Skip to content
Branch: master
Find file Copy path
Find file Copy path
Fetching contributors…
Cannot retrieve contributors at this time
1600 lines (1269 sloc) 62.2 KB

GnuPG Details


This is the DETAILS file for GnuPG which specifies some internals and parts of the external API for GPG and GPGSM.

Format of the colon listings

The format is a based on colon separated record, each recods starts with a tag string and extends to the end of the line. Here is an example:

$ gpg --with-colons --list-keys \
      --with-fingerprint --with-fingerprint
uid:f::::::::Werner Koch <>:
uid:f::::::::Werner Koch <>:

Note that new version of GnuPG or the use of certain options may add new fields to the output. Parsers should not assume a limit on the number of fields per line. Some fields are not yet used or only used with certain record types; parsers should ignore fields they are not aware of. New versions of GnuPG or the use of certain options may add new types of records as well. Parsers should ignore any record whose type they do not recognize for forward-compatibility.

The double --with-fingerprint prints the fingerprint for the subkeys too. Old versions of gpg used a slightly different format and required the use of the option --fixed-list-mode to conform to the format described here.

Description of the fields

Field 1 - Type of record

Public key
X.509 certificate
X.509 certificate and private key available
Subkey (secondary key)
Secret key
Secret subkey (secondary key)
User id
User attribute (same as user id except for field 10).
Revocation signature
Revocation signature (standalone) [since 2.2.9]
Fingerprint (fingerprint is in field 10)
Public key data [*]
Revocation key
TOFU statistics [*]
Trust database information [*]
Signature subpacket [*]
Configuration data [*]

Records marked with an asterisk are described at *Special fields.

Field 2 - Validity

This is a letter describing the computed validity of a key. Currently this is a single letter, but be prepared that additional information may follow in some future versions. Note that GnuPG < 2.1 does not set this field for secret key listings.

Unknown (this key is new to the system)
The key is invalid (e.g. due to a missing self-signature)
The key has been disabled

(deprecated - use the ‘D’ in field 12 instead)

The key has been revoked
The key has expired
Unknown validity (i.e. no value assigned)
Undefined validity. ‘-’ and ‘q’ may safely be treated as the same value for most purposes
The key is not valid
The key is marginal valid.
The key is fully valid
The key is ultimately valid. This often means that the secret key is available, but any key may be marked as ultimately valid.
The key has a well known private part.
The key has special validity. This means that it might be self-signed and expected to be used in the STEED system.

If the validity information is given for a UID or UAT record, it describes the validity calculated based on this user ID. If given for a key record it describes the validity taken from the best rated user ID.

For X.509 certificates a ‘u’ is used for a trusted root certificate (i.e. for the trust anchor) and an ‘f’ for all other valid certificates.

In “sig” records, this field may have one of these values as first character:

Signature is good.
Signature is bad.
No public key to verify signature or public key is not usable.
Other error verifying a signature

More values may be added later. The field may also be empty if gpg has been invoked in a non-checking mode (–list-sigs) or in a fast checking mode. Since 2.2.7 ‘?’ will also be printed by the command –list-sigs if the key is not in the local keyring.

Field 3 - Key length

The length of key in bits.

Field 4 - Public key algorithm

The values here are those from the OpenPGP specs or if they are greater than 255 the algorithm ids as used by Libgcrypt.

Field 5 - KeyID

This is the 64 bit keyid as specified by OpenPGP and the last 64 bit of the SHA-1 fingerprint of an X.509 certifciate.

Field 6 - Creation date

The creation date of the key is given in UTC. For UID and UAT records, this is used for the self-signature date. Note that the date is usually printed in seconds since epoch, however, we are migrating to an ISO 8601 format (e.g. “19660205T091500”). This is currently only relevant for X.509. A simple way to detect the new format is to scan for the ‘T’. Note that old versions of gpg without using the --fixed-list-mode option used a “yyyy-mm-tt” format.

Field 7 - Expiration date

Key or UID/UAT expiration date or empty if it does not expire.

Field 8 - Certificate S/N, UID hash, trust signature info

Used for serial number in crt records. For UID and UAT records, this is a hash of the user ID contents used to represent that exact user ID. For trust signatures, this is the trust depth separated by the trust value by a space.

Field 9 - Ownertrust

This is only used on primary keys. This is a single letter, but be prepared that additional information may follow in future versions. For trust signatures with a regular expression, this is the regular expression value, quoted as in field 10.

Field 10 - User-ID

The value is quoted like a C string to avoid control characters (the colon is quoted \x3a). For a “pub” record this field is not used on –fixed-list-mode. A UAT record puts the attribute subpacket count here, a space, and then the total attribute subpacket size. In gpgsm the issuer name comes here. A FPR record stores the fingerprint here. The fingerprint of a revocation key is stored here.

Field 11 - Signature class

Signature class as per RFC-4880. This is a 2 digit hexnumber followed by either the letter ‘x’ for an exportable signature or the letter ‘l’ for a local-only signature. The class byte of an revocation key is also given here, by a 2 digit hexnumber and optionally followed by the letter ‘s’ for the “sensitive” flag. This field is not used for X.509.

“rev” and “rvs” may be followed by a comma and a 2 digit hexnumber with the revocation reason.

Field 12 - Key capabilities

The defined capabilities are:

Unknown capability

A key may have any combination of them in any order. In addition to these letters, the primary key has uppercase versions of the letters to denote the usable capabilities of the entire key, and a potential letter ‘D’ to indicate a disabled key.

Field 13 - Issuer certificate fingerprint or other info

Used in FPR records for S/MIME keys to store the fingerprint of the issuer certificate. This is useful to build the certificate path based on certificates stored in the local key database it is only filled if the issuer certificate is available. The root has been reached if this is the same string as the fingerprint. The advantage of using this value is that it is guaranteed to have been built by the same lookup algorithm as gpgsm uses.

For “uid” records this field lists the preferences in the same way gpg’s –edit-key menu does.

For “sig”, “rev” and “rvs” records, this is the fingerprint of the key that issued the signature. Note that this may only be filled if the signature verified correctly. Note also that for various technical reasons, this fingerprint is only available if –no-sig-cache is used. Since 2.2.7 this field will also be set if the key is missing but the signature carries an issuer fingerprint as meta data.

Field 14 - Flag field

Flag field used in the –edit menu output

Field 15 - S/N of a token

Used in sec/ssb to print the serial number of a token (internal protect mode 1002) or a ‘#’ if that key is a simple stub (internal protect mode 1001). If the option –with-secret is used and a secret key is available for the public key, a ‘+’ indicates this.

Field 16 - Hash algorithm

For sig records, this is the used hash algorithm. For example: 2 = SHA-1, 8 = SHA-256.

Field 17 - Curve name

For pub, sub, sec, and ssb records this field is used for the ECC curve name.

Field 18 - Compliance flags

Space separated list of asserted compliance modes and screening result for this key.

Valid values are:

The key is compliant with RFC4880bis
The key is compliant with compliance mode “de-vs”.
Screening hit on the ROCA vulnerability.

Field 19 - Last update

The timestamp of the last update of a key or user ID. The update time of a key is defined a lookup of the key via its unique identifier (fingerprint); the field is empty if not known. The update time of a user ID is defined by a lookup of the key using a trusted mapping from mail address to key.

Field 20 - Origin

The origin of the key or the user ID. This is an integer optionally followed by a space and an URL. This goes along with the previous field. The URL is quoted in C style.

Field 21 - Comment

This is currently only used in “rev” and “rvs” records to carry the the comment field of the recocation reason. The value is quoted in C style.

Special fields

PKD - Public key data

If field 1 has the tag “pkd”, a listing looks like this:

pkd:0:1024:B665B1435F4C2 .... FF26ABB:
    !  !   !-- the value
    !  !------ for information number of bits in the value
    !--------- index (eg. DSA goes from 0 to 3: p,q,g,y)

TFS - TOFU statistics

This field may follows a UID record to convey information about the TOFU database. The information is similar to a TOFU_STATS status line.

Field 2
tfs record version (must be 1)
Field 3
validity - A number with validity code.
Field 4
signcount - The number of signatures seen.
Field 5
encrcount - The number of encryptions done.
Field 6
policy - A string with the policy
Field 7
signture-first-seen - a timestamp or 0 if not known.
Field 8
signature-most-recent-seen - a timestamp or 0 if not known.
Field 9
encryption-first-done - a timestamp or 0 if not known.
Field 10
encryption-most-recent-done - a timestamp or 0 if not known.

TRU - Trust database information

Example for a “tru” trust base record:

Field 2
Reason for staleness of trust. If this field is empty, then the trustdb is not stale. This field may have multiple flags in it:
Trustdb is old
Trustdb was built with a different trust model than the one we are using now.
Field 3
Trust model
Classic trust model, as used in PGP 2.x.
PGP trust model, as used in PGP 6 and later. This is the same as the classic trust model, except for the addition of trust signatures.

GnuPG before version 1.4 used the classic trust model by default. GnuPG 1.4 and later uses the PGP trust model by default.

Field 4
Date trustdb was created in seconds since Epoch.
Field 5
Date trustdb will expire in seconds since Epoch.
Field 6
Number of marginally trusted users to introduce a new key signer (gpg’s option –marginals-needed).
Field 7
Number of completely trusted users to introduce a new key signer. (gpg’s option –completes-needed)
Field 8
Maximum depth of a certification chain. (gpg’s option –max-cert-depth)

SPK - Signature subpacket records

Field 2
Subpacket number as per RFC-4880 and later.
Field 3
Flags in hex. Currently the only two bits assigned are 1, to indicate that the subpacket came from the hashed part of the signature, and 2, to indicate the subpacket was marked critical.
Field 4
Length of the subpacket. Note that this is the length of the subpacket, and not the length of field 5 below. Due to the need for %-encoding, the length of field 5 may be up to 3x this value.
Field 5
The subpacket data. Printable ASCII is shown as ASCII, but other values are rendered as %XX where XX is the hex value for the byte.

CFG - Configuration data

–list-config outputs information about the GnuPG configuration for the benefit of frontends or other programs that call GnuPG. There are several list-config items, all colon delimited like the rest of the –with-colons output. The first field is always “cfg” to indicate configuration information. The second field is one of (with examples):

The third field contains the version of GnuPG.
The third field contains the public key algorithms this version of GnuPG supports, separated by semicolons. The algorithm numbers are as specified in RFC-4880. Note that in contrast to the –status-fd interface these are not the Libgcrypt identifiers. Using pubkeyname prints names instead of numbers.
The third field contains the symmetric ciphers this version of GnuPG supports, separated by semicolons. The cipher numbers are as specified in RFC-4880. Using ciphername prints names instead of numbers.
The third field contains the digest (hash) algorithms this version of GnuPG supports, separated by semicolons. The digest numbers are as specified in RFC-4880. Using digestname prints names instead of numbers.
The third field contains the compression algorithms this version of GnuPG supports, separated by semicolons. The algorithm numbers are as specified in RFC-4880.
The third field contains the name of the group, and the fourth field contains the values that the group expands to, separated by semicolons.

For example, a group of:

group mynames = paige 0x12345678 joe patti

would result in:

The third field contains the curve names this version of GnuPG supports, separated by semicolons. Using curveoid prints OIDs instead of numbers.

Format of the –status-fd output

Every line is prefixed with “[GNUPG:] “, followed by a keyword with the type of the status line and some arguments depending on the type (maybe none); an application should always be willing to ignore unknown keywords that may be emitted by future versions of GnuPG. Also, new versions of GnuPG may add arguments to existing keywords. Any additional arguments should be ignored for forward-compatibility.

General status codes

NEWSIG [<signers_uid>]

Is issued right before a signature verification starts. This is useful to define a context for parsing ERROR status messages. If SIGNERS_UID is given and is not “-” this is the percent-escaped value of the OpenPGP Signer’s User ID signature sub-packet.

GOODSIG <long_keyid_or_fpr> <username>

The signature with the keyid is good. For each signature only one of the codes GOODSIG, BADSIG, EXPSIG, EXPKEYSIG, REVKEYSIG or ERRSIG will be emitted. In the past they were used as a marker for a new signature; new code should use the NEWSIG status instead. The username is the primary one encoded in UTF-8 and %XX escaped. The fingerprint may be used instead of the long keyid if it is available. This is the case with CMS and might eventually also be available for OpenPGP.

EXPSIG <long_keyid_or_fpr> <username>

The signature with the keyid is good, but the signature is expired. The username is the primary one encoded in UTF-8 and %XX escaped. The fingerprint may be used instead of the long keyid if it is available. This is the case with CMS and might eventually also be available for OpenPGP.

EXPKEYSIG <long_keyid_or_fpr> <username>

The signature with the keyid is good, but the signature was made by an expired key. The username is the primary one encoded in UTF-8 and %XX escaped. The fingerprint may be used instead of the long keyid if it is available. This is the case with CMS and might eventually also be available for OpenPGP.

REVKEYSIG <long_keyid_or_fpr> <username>

The signature with the keyid is good, but the signature was made by a revoked key. The username is the primary one encoded in UTF-8 and %XX escaped. The fingerprint may be used instead of the long keyid if it is available. This is the case with CMS and might eventually also beñ available for OpenPGP.

BADSIG <long_keyid_or_fpr> <username>

The signature with the keyid has not been verified okay. The username is the primary one encoded in UTF-8 and %XX escaped. The fingerprint may be used instead of the long keyid if it is available. This is the case with CMS and might eventually also be available for OpenPGP.

ERRSIG <keyid> <pkalgo> <hashalgo> <sig_class> <time> <rc> <fpr>

It was not possible to check the signature. This may be caused by a missing public key or an unsupported algorithm. A RC of 4 indicates unknown algorithm, a 9 indicates a missing public key. The other fields give more information about this signature. sig_class is a 2 byte hex-value. The fingerprint may be used instead of the long_keyid_or_fpr if it is available. This is the case with gpgsm and might eventually also be available for OpenPGP. The ERRSIG line has FPR filed which is only available since 2.2.7; that FPR may either be missing or - if the signature has no fingerprint as meta data.

Note, that TIME may either be the number of seconds since Epoch or an ISO 8601 string. The latter can be detected by the presence of the letter ‘T’.


The args are:

  • <fingerprint_in_hex>
  • <sig_creation_date>
  • <sig-timestamp>
  • <expire-timestamp>
  • <sig-version>
  • <reserved>
  • <pubkey-algo>
  • <hash-algo>
  • <sig-class>
  • [ <primary-key-fpr> ]

This status indicates that the signature is cryptographically valid. This is similar to GOODSIG, EXPSIG, EXPKEYSIG, or REVKEYSIG (depending on the date and the state of the signature and signing key) but has the fingerprint as the argument. Multiple status lines (VALIDSIG and the other appropriate *SIG status) are emitted for a valid signature. All arguments here are on one long line. sig-timestamp is the signature creation time in seconds after the epoch. expire-timestamp is the signature expiration time in seconds after the epoch (zero means “does not expire”). sig-version, pubkey-algo, hash-algo, and sig-class (a 2-byte hex value) are all straight from the signature packet. PRIMARY-KEY-FPR is the fingerprint of the primary key or identical to the first argument. This is useful to get back to the primary key without running gpg again for this purpose.

The primary-key-fpr parameter is used for OpenPGP and not available for CMS signatures. The sig-version as well as the sig class is not defined for CMS and currently set to 0 and 00.

Note, that *-TIMESTAMP may either be a number of seconds since Epoch or an ISO 8601 string which can be detected by the presence of the letter ‘T’.

SIG_ID <radix64_string> <sig_creation_date> <sig-timestamp>

This is emitted only for signatures of class 0 or 1 which have been verified okay. The string is a signature id and may be used in applications to detect replay attacks of signed messages. Note that only DLP algorithms give unique ids - others may yield duplicated ones when they have been created in the same second.

Note, that SIG-TIMESTAMP may either be a number of seconds since Epoch or an ISO 8601 string which can be detected by the presence of the letter ‘T’.

ENC_TO <long_keyid> <keytype> <keylength>

The message is encrypted to this LONG_KEYID. KEYTYPE is the numerical value of the public key algorithm or 0 if it is not known, KEYLENGTH is the length of the key or 0 if it is not known (which is currently always the case). Gpg prints this line always; Gpgsm only if it knows the certificate.


Mark the start of the actual decryption process. This is also emitted when in –list-only mode.


Mark the end of the actual decryption process. This are also emitted when in –list-only mode.

DECRYPTION_KEY <fpr> <fpr2> <otrust>

This line is emitted when a public key decryption succeeded in providing a session key. <fpr> is the hexified fingerprint of the actual key used for decryption. <fpr2> is the fingerprint of the primary key. <otrust> is the letter with the ownertrust; this is in general a ‘u’ which stands for ultimately trusted.

DECRYPTION_INFO <mdc_method> <sym_algo> [<aead_algo>]

Print information about the symmetric encryption algorithm and the MDC method. This will be emitted even if the decryption fails. For an AEAD algorithm AEAD_ALGO is not 0. GPGSM currently does not print such a status.


The symmetric decryption failed - one reason could be a wrong passphrase for a symmetrical encrypted message.


The decryption process succeeded. This means, that either the correct secret key has been used or the correct passphrase for a symmetric encrypted message was given. The program itself may return an errorcode because it may not be possible to verify a signature for some reasons.

SESSION_KEY <algo>:<hexdigits>

The session key used to decrypt the message. This message will only be emitted if the option –show-session-key is used. The format is suitable to be passed as value for the option –override-session-key. It is not an indication that the decryption will or has succeeded.

BEGIN_ENCRYPTION <mdc_method> <sym_algo> [<aead_algo>]

Mark the start of the actual encryption process. MDC_METHOD shall be 0 if an AEAD_ALGO is not 0. Users should however ignore MDC_METHOD if AEAD_ALGO is not 0.


Mark the end of the actual encryption process.

FILE_START <what> <filename>

Start processing a file <filename>. <what> indicates the performed operation:



Marks the end of a file processing which has been started by FILE_START.


Mark the start of the actual signing process. This may be used as an indication that all requested secret keys are ready for use.

ALREADY_SIGNED <long-keyid>

Warning: This is experimental and might be removed at any time.

SIG_CREATED <type> <pk_algo> <hash_algo> <class> <timestamp> <keyfpr>

A signature has been created using these parameters. Values for type <type> are:


(only the first character should be checked)

<class> are 2 hex digits with the OpenPGP signature class.

Note, that TIMESTAMP may either be a number of seconds since Epoch or an ISO 8601 string which can be detected by the presence of the letter ‘T’.


There are actually three related status codes to convey notation data:

  • NOTATION_NAME <name>
  • NOTATION_FLAGS <critical> <human_readable>
  • NOTATION_DATA <string>

<name> and <string> are %XX escaped. The data may be split among several NOTATION_DATA lines. NOTATION_FLAGS is emitted after NOTATION_NAME and gives the critical and human readable flags; the flag values are either 0 or 1.

POLICY_URL <string>

Note that URL in <string> is %XX escaped.

PLAINTEXT <format> <timestamp> <filename>

This indicates the format of the plaintext that is about to be written. The format is a 1 byte hex code that shows the format of the plaintext: 62 (‘b’) is binary data, 74 (‘t’) is text data with no character set specified, and 75 (‘u’) is text data encoded in the UTF-8 character set. The timestamp is in seconds since the epoch. If a filename is available it gets printed as the third argument, percent-escaped as usual.


This indicates the length of the plaintext that is about to be written. Note that if the plaintext packet has partial length encoding it is not possible to know the length ahead of time. In that case, this status tag does not appear. The length is only exact for binary formats; other formats (‘t’, ‘u’) may do post processing like line ending conversion so that the actual number of bytes written may be differ.

ATTRIBUTE <arguments>

The list or arguments are:

  • <fpr>
  • <octets>
  • <type>
  • <index>
  • <count>
  • <timestamp>
  • <expiredate>
  • <flags>

This is one long line issued for each attribute subpacket when an attribute packet is seen during key listing. <fpr> is the fingerprint of the key. <octets> is the length of the attribute subpacket. <type> is the attribute type (e.g. 1 for an image). <index> and <count> indicate that this is the N-th indexed subpacket of count total subpackets in this attribute packet. <timestamp> and <expiredate> are from the self-signature on the attribute packet. If the attribute packet does not have a valid self-signature, then the timestamp is 0. <flags> are a bitwise OR of:

this attribute packet is a primary uid
this attribute packet is revoked
this attribute packet is expired

SIG_SUBPACKET <type> <flags> <len> <data>

This indicates that a signature subpacket was seen. The format is the same as the “spk” record above.


Indicates that the current encryption operation was in compliance with the given set of modes for all recipients. “flags” is a space separated list of numerical flags, see “Field 18 - Compliance flags” above.


Indicates that the current decryption operation is in compliance with the given set of modes. “flags” is a space separated list of numerical flags, see “Field 18 - Compliance flags” above.


Indicates that the current signature verification operation is in compliance with the given set of modes. “flags” is a space separated list of numerical flags, see “Field 18 - Compliance flags” above.

Key related


The two similar status codes:

  • INV_RECP <reason> <requested_recipient>
  • INV_SGNR <reason> <requested_sender>

are issued for each unusable recipient/sender. The reasons codes currently in use are:

No specific reason given
Not Found
Ambiguous specification
Wrong key usage
Key revoked
Key expired
No CRL known
CRL too old
Policy mismatch
Not a secret key
Key not trusted
Missing certificate
Missing issuer certificate
Key disabled
Syntax error in specification

If no specific reason was given a previously emitted status code KEY_CONSIDERED may be used to analyzed the problem.

Note that for historical reasons the INV_RECP status is also used for gpgsm’s SIGNER command where it relates to signer’s of course. Newer GnuPG versions are using INV_SGNR; applications should ignore the INV_RECP during the sender’s command processing once they have seen an INV_SGNR. Different codes are used so that they can be distinguish while doing an encrypt+sign operation.

NO_RECP <reserved>

Issued if no recipients are usable.

NO_SGNR <reserved>

Issued if no senders are usable.

KEY_CONSIDERED <fpr> <flags>

Issued to explian the lookup of a key. FPR is the hexified fingerprint of the primary key. The bit values for FLAGS are:

The key has not been selected.
All subkeys of the key are expired or have been revoked.

KEYEXPIRED <expire-timestamp>

The key has expired. expire-timestamp is the expiration time in seconds since Epoch. This status line is not very useful because it will also be emitted for expired subkeys even if this subkey is not used. To check whether a key used to sign a message has expired, the EXPKEYSIG status line is to be used.

Note, that the TIMESTAMP may either be a number of seconds since Epoch or an ISO 8601 string which can be detected by the presence of the letter ‘T’.


The used key has been revoked by its owner. No arguments yet.

NO_PUBKEY <long keyid>

The public key is not available. Note the arg should in general not be used because it is better to take it from the ERRSIG status line which is printed right before this one.

NO_SECKEY <long keyid>

The secret key is not available

KEY_CREATED <type> <fingerprint> [<handle>]

A key has been created. Values for <type> are:

primary and subkey

The fingerprint is one of the primary key for type B and P and the one of the subkey for S. Handle is an arbitrary non-whitespace string used to match key parameters from batch key creation run.

KEY_NOT_CREATED [<handle>]

The key from batch run has not been created due to errors.


These are several similar status codes:

  • TRUST_UNDEFINED <error_token>
  • TRUST_NEVER <error_token>
  • TRUST_MARGINAL [0 [<validation_model>]]
  • TRUST_FULLY [0 [<validation_model>]]
  • TRUST_ULTIMATE [0 [<validation_model>]]

For good signatures one of these status lines are emitted to indicate the validity of the key used to create the signature. The error token values are currently only emitted by gpgsm.

VALIDATION_MODEL describes the algorithm used to check the validity of the key. The defaults are the standard Web of Trust model for gpg and the standard X.509 model for gpgsm. The defined values are

The standard PGP WoT.
The standard X.509 model.
The chain model.
The STEED model.
The TOFU model

Note that the term TRUST_ in the status names is used for historic reasons; we now speak of validity.

TOFU_USER <fingerprint_in_hex> <mbox>

This status identifies the key and the userid for all following Tofu information. The fingerprint is the fingerprint of the primary key and the mbox is in general the addr-spec part of the userid encoded in UTF-8 and percent escaped. The fingerprint is identical for all TOFU_USER lines up to a NEWSIG line.


Statistics for the current user id.

The <MANY_ARGS> are the usual space delimited arguments. Here we have too many of them to fit on one printed line and thus they are given on 3 printed lines:

<summary> <sign-count> <encryption-count>
[<policy> [<tm1> <tm2> <tm3> <tm4>
[<validity> [<sign-days> <encrypt-days>]]]]

Values for SUMMARY are:

attention, an interaction with the user is required (conflict)
key with no verification/encryption history
key with little history
key with enough history for basic trust
key with a lot of history

Values for POLICY are:

No Policy set
Policy is “auto”
Policy is “good”
Policy is “bad”
Policy is “ask”
Policy is “unknown” (TOFU information does not contribute to the key’s validity)

TM1 is the time the first message was verified. TM2 is the time the most recent message was verified. TM3 is the time the first message was encrypted. TM4 is the most recent encryption. All may either be seconds since Epoch or an ISO time string (yyyymmddThhmmss).

VALIDITY is the same as SUMMARY with the exception that VALIDITY doesn’t reflect whether the key needs attention. That is it never takes on value 0. Instead, if there is a conflict, VALIDITY still reflects the key’s validity (values: 1-4).

SUMMARY values use the euclidean distance (m = sqrt(a² + b²)) rather then the sum of the magnitudes (m = a + b) to ensure a balance between verified signatures and encrypted messages.

Values are calculated based on the number of days where a key was used for verifying a signature or to encrypt to it. The ranges for the values are:

signature_days + encryption_days == 0
1 <= sqrt(signature_days² + encryption_days²) < 8
8 <= sqrt(signature_days² + encryption_days²) < 42
sqrt(signature_days² + encryption_days²) >= 42

SIGN-COUNT and ENCRYPTION-COUNT are the number of messages that we have seen that have been signed by this key / encryption to this key.

SIGN-DAYS and ENCRYPTION-DAYS are similar, but the number of days (in UTC) on which we have seen messages signed by this key / encrypted to this key.

TOFU_STATS_SHORT <long_string>

Information about the TOFU binding for the signature. Example: “15 signatures verified. 10 messages encrypted”

TOFU_STATS_LONG <long_string>

Information about the TOFU binding for the signature in verbose format. The LONG_STRING is percent escaped. Example: ‘Verified 9 messages signed by “Werner Koch (dist sig)” in the past 3 minutes, 40 seconds. The most recent message was verified 4 seconds ago.’


This is one of:

  • PKA_TRUST_GOOD <addr-spec>
  • PKA_TRUST_BAD <addr-spec>

Depending on the outcome of the PKA check one of the above status codes is emitted in addition to a TRUST_* status.

Remote control


These status line are used with –command-fd for interactive control of the process.

USERID_HINT <long main keyid> <string>

Give a hint about the user ID for a certain keyID.

NEED_PASSPHRASE <long keyid> <long main keyid> <keytype> <keylength>

Issued whenever a passphrase is needed. KEYTYPE is the numerical value of the public key algorithm or 0 if this is not applicable, KEYLENGTH is the length of the key or 0 if it is not known (this is currently always the case).

NEED_PASSPHRASE_SYM <cipher_algo> <s2k_mode> <s2k_hash>

Issued whenever a passphrase for symmetric encryption is needed.

NEED_PASSPHRASE_PIN <card_type> <chvno> [<serialno>]

Issued whenever a PIN is requested to unlock a card.


No passphrase was supplied. An application which encounters this message may want to stop parsing immediately because the next message will probably be a BAD_PASSPHRASE. However, if the application is a wrapper around the key edit menu functionality it might not make sense to stop parsing but simply ignoring the following BAD_PASSPHRASE.

BAD_PASSPHRASE <long keyid>

The supplied passphrase was wrong or not given. In the latter case you may have seen a MISSING_PASSPHRASE.


The supplied passphrase was good and the secret key material is therefore usable.


IMPORT_CHECK <long keyid> <fingerprint> <user ID>

This status is emitted in interactive mode right before the “import.okay” prompt.

IMPORTED <long keyid> <username>

The keyid and name of the signature just imported

IMPORT_OK <reason> [<fingerprint>]

The key with the primary key’s FINGERPRINT has been imported. REASON flags are:

Not actually changed
Entirely new key.
New user IDs
New signatures
New subkeys
Contains private key.

The flags may be ORed.

IMPORT_PROBLEM <reason> [<fingerprint>]

Issued for each import failure. Reason codes are:

No specific reason given.
Invalid Certificate.
Issuer Certificate missing.
Certificate Chain too long.
Error storing certificate.


Final statistics on import process (this is one long line). The args are a list of unsigned numbers separated by white space:

  • <count>
  • <no_user_id>
  • <imported>
  • always 0 (formerly used for the number of RSA keys)
  • <unchanged>
  • <n_uids>
  • <n_subk>
  • <n_sigs>
  • <n_revoc>
  • <sec_read>
  • <sec_imported>
  • <sec_dups>
  • <skipped_new_keys>
  • <not_imported>
  • <skipped_v3_keys>

EXPORTED <fingerprint>

The key with <fingerprint> has been exported. The fingerprint is the fingerprint of the primary key even if the primary key has been replaced by a stub key during secret key export.


Final statistics on export process (this is one long line). The args are a list of unsigned numbers separated by white space:

  • <count>
  • <secret_count>
  • <exported>

Smartcard related

CARDCTRL <what> [<serialno>]

This is used to control smartcard operations. Defined values for WHAT are:

Request insertion of a card. Serialnumber may be given to request a specific card. Used by gpg 1.4 w/o scdaemon
Request removal of a card. Used by gpg 1.4 w/o scdaemon.
Card with serialnumber detected
No card available
No card reader available
No card support available
Card is in termination state

SC_OP_FAILURE [<code>]

An operation on a smartcard definitely failed. Currently there is no indication of the actual error code, but application should be prepared to later accept more arguments. Defined values for <code> are:

unspecified error (identically to a missing CODE)
bad PIN


A smart card operation succeeded. This status is only printed for certain operation and is mostly useful to check whether a PIN change really worked.

Miscellaneous status codes

NODATA <what>

No data has been found. Codes for WHAT are:

No armored data.
Expected a packet but did not found one.
Invalid packet found, this may indicate a non OpenPGP message.
Signature expected but not found

You may see more than one of these status lines.


Unexpected data has been encountered. Codes for WHAT are:

Not further specified
Corrupted message structure


The output was truncated to MAXNO items. This status code is issued for certain external requests.

ERROR <error location> <error code> [<more>]

This is a generic error status message, it might be followed by error location specific data. <error code> and <error_location> should not contain spaces. The error code is a either a string commencing with a letter or such a string prefixed with a numerical error code and an underscore; e.g.: “151011327_EOF”.

WARNING <location> <error code> [<text>]

This is a generic warning status message, it might be followed by error location specific data. <location> and <error code> may not contain spaces. The <location> may be used to indicate a class of warnings. The error code is a either a string commencing with a letter or such a string prefixed with a numerical error code and an underscore; e.g.: “151011327_EOF”.

NOTE <location> <error code> [<text>]

This is a generic info status message the same syntax as for WARNING messages is used.

SUCCESS [<location>]

Positive confirmation that an operation succeeded. It is used similar to ISO-C’s EXIT_SUCCESS. <location> is optional but if given should not contain spaces. Used only with a few commands.

FAILURE <location> <error_code>

This is the counterpart to SUCCESS and used to indicate a program failure. It is used similar to ISO-C’s EXIT_FAILURE but allows conveying more information, in particular a gpg-error error code. That numerical error code may optionally have a suffix made of an underscore and a string with an error symbol like “151011327_EOF”. A dash may be used instead of <location>.


The ASCII armor is corrupted. No arguments yet.

DELETE_PROBLEM <reason_code>

Deleting a key failed. Reason codes are:

No such key
Must delete secret key first
Ambiguous specification
Key is stored on a smartcard.

PROGRESS <what> <char> <cur> <total> [<units>]

Used by the primegen and public key functions to indicate progress. <char> is the character displayed with no –status-fd enabled, with the linefeed replaced by an ‘X’. <cur> is the current amount done and <total> is amount to be done; a <total> of 0 indicates that the total amount is not known. Both are non-negative integers. The condition

      TOTAL && CUR == TOTAL

may be used to detect the end of an operation.

Well known values for <what> are:

DSA key generation
Elgamal key generation
Prime generation
Waiting for new entropy in the RNG
Generic tick without any special meaning - useful for letting clients know that the server is still working.
A gpg-agent was started because it is not running as a daemon.
Send by the agent and gpgsm while learing the data of a smartcard.
A smartcard is still working

When <what> refers to a file path, it may be truncated.

<units> is sometimes used to describe the units for <current> and <total>. For example “B”, “KiB”, or “MiB”.

BACKUP_KEY_CREATED <fingerprint> <fname>

A backup of a key identified by <fingerprint> has been writte to the file <fname>; <fname> is percent-escaped.


<name> is a percent-plus escaped filename describing the mountpoint for the current operation (e.g. used by “g13 –mount”). This may either be the specified mountpoint or one randomly chosen by g13.

PINENTRY_LAUNCHED <pid>[:<extra>]

This status line is emitted by gpg to notify a client that a Pinentry has been launched. <pid> is the PID of the Pinentry. It may be used to display a hint to the user but can’t be used to synchronize with Pinentry. Note that there is also an Assuan inquiry line with the same name used internally or, if enabled, send to the client instead of this status line. Such an inquiry may be used to sync with Pinentry

Obsolete status codes


Removed on 2011-02-04. This is deprecated in favor of KEYEXPIRED.


Obsolete. This status message used to be emitted for requests to use the IDEA or RSA algorithms. It has been dropped from GnuPG 2.1 after the respective patents expired.


These were used for the ancient shared memory based co-processing.


Used to issued by the experimental pipemode.

Inter-component codes

Status codes are also used between the components of the GnuPG system via the Assuan S lines. Some of them are documented here:

PUBKEY_INFO <n> <ubid>

The type of the public key in the following D-lines or communicated via a pipe. <n> is the value of enum pubkey_types and <ubid> the Unique Blob ID (UBID) which is the fingerprint of the primary key truncated to 20 octets and formatted in hex. Note that the keyboxd SEARCH command can be used to lookup the public key using the <ubid> prefixed with a caret (^).

Format of the –attribute-fd output

When –attribute-fd is set, during key listings (–list-keys, –list-secret-keys) GnuPG dumps each attribute packet to the file descriptor specified. –attribute-fd is intended for use with –status-fd as part of the required information is carried on the ATTRIBUTE status tag (see above).

The contents of the attribute data is specified by RFC 4880. For convenience, here is the Photo ID format, as it is currently the only attribute defined:

Byte 0-1
The length of the image header. Due to a historical accident (i.e. oops!) back in the NAI PGP days, this is a little-endian number. Currently 16 (0x10 0x00).
Byte 2
The image header version. Currently 0x01.
Byte 3
Encoding format. 0x01 == JPEG.
Byte 4-15
Reserved, and currently unused.

All other data after this header is raw image (JPEG) data.

Layout of the TrustDB

The TrustDB is built from fixed length records, where the first byte describes the record type. All numeric values are stored in network byte order. The length of each record is 40 bytes. The first record of the DB is always of type 1 and this is the only record of this type.

The record types: directory(2), key(3), uid(4), pref(5), sigrec(6), and shadow directory(8) are not anymore used by version 2 of the TrustDB.

Record type 0

Unused record or deleted, can be reused for any purpose. Such records should in general not exist because deleted records are of type 254 and kept in a linked list.

Version info (RECTYPE_VER, 1)

Version information for this TrustDB. This is always the first record of the DB and the only one of this type.

1 u8
Record type (value: 1).
3 byte
Magic value (“gpg”)
1 u8
TrustDB version (value: 2).
1 u8
marginals. How many marginal trusted keys are required.
1 u8
completes. How many completely trusted keys are required.
1 u8
max_cert_depth. How deep is the WoT evaluated. Along with marginals and completes, this value is used to check whether the cached validity value from a [FIXME dir] record can be used.
1 u8
1 u8
2 byte
Not used
1 u32
created. Timestamp of trustdb creation.
1 u32
nextcheck. Timestamp of last modification which may affect the validity of keys in the trustdb. This value is checked against the validity timestamp in the dir records.
1 u32
reserved. Not used.
1 u32
reserved2. Not used.
1 u32
firstfree. Number of the record with the head record of the RECTYPE_FREE linked list.
1 u32
reserved3. Not used.
1 u32
trusthashtbl. Record number of the trusthashtable.

Hash table (RECTYPE_HTBL, 10)

Due to the fact that we use fingerprints to lookup keys, we can implement quick access by some simple hash methods, and avoid the overhead of gdbm. A property of fingerprints is that they can be used directly as hash values. What we use is a dynamic multilevel architecture, which combines hash tables, record lists, and linked lists.

This record is a hash table of 256 entries with the property that all these records are stored consecutively to make one big table. The hash value is simple the 1st, 2nd, … byte of the fingerprint (depending on the indirection level).

1 u8
Record type (value: 10).
1 u8
n u32
recnum. A table with the hash table items fitting into this record. n depends on the record length: $n=(reclen-2)/4$ which yields 9 for oure current record length of 40 bytes.

The total number of hash table records to form the table is: $m=(256+n-1)/n$. This is 29 for our record length of 40.

To look up a key we use the first byte of the fingerprint to get the recnum from this hash table and then look up the addressed record:

  • If that record is another hash table, we use 2nd byte to index that hash table and so on;
  • if that record is a hash list, we walk all entries until we find a matching one; or
  • if that record is a key record, we compare the fingerprint to decide whether it is the requested key;

Hash list (RECTYPE_HLST, 11)

See hash table above on how it is used. It may also be used for other purposes.

1 u8
Record type (value: 11).
1 u8
1 u32
next. Record number of the next hash list record or 0 if none.
n u32
rnum. Array with record numbers to values. With $n=(reclen-5)/5$ and our record length of 40, n is 7.

Trust record (RECTYPE_TRUST, 12)

1 u8
Record type (value: 12).
1 u8
20 byte
1 u8
1 u8
1 u8
1 byte
Not used.
1 u32
10 byte
Not used.

Validity record (RECTYPE_VALID, 13)

1 u8
Record type (value: 13).
1 u8
20 byte
1 u8
1 u32
1 u8
1 u8
11 byte
Not used.

Free record (RECTYPE_FREE, 254)

All these records form a linked list of unused records in the TrustDB.

1 u8
Record type (value: 254)
1 u8
1 u32
next. Record number of the next rcord of this type. The record number to the head of this linked list is stored in the version info record.

Database scheme for the TOFU info

-- The VERSION table holds the version of our TOFU data structures.
CREATE TABLE version (
  version integer -- As of now this is always 1

-- The BINDINGS table associates mail addresses with keys.
CREATE TABLE bindings (
  oid integer primary key autoincrement,
  fingerprint text, -- The key's fingerprint in hex
  email text,       -- The normalized mail address destilled from user_id
  user_id text,     -- The unmodified user id
  time integer,     -- The time this binding was first observed.
  policy boolean check
       (policy in (1, 2, 3, 4, 5)), -- The trust policy with the values:
                                    --   1 := Auto
                                    --   2 := Good
                                    --   3 := Unknown
                                    --   4 := Bad
                                    --   5 := Ask
  conflict string,  -- NULL or a hex formatted fingerprint.
  unique (fingerprint, email)

CREATE INDEX bindings_fingerprint_email on bindings (fingerprint, email);
CREATE INDEX bindings_email on bindings (email);

-- The SIGNATURES table records all data signatures we verified
CREATE TABLE signatures (
  binding integer not null, -- Link to bindings table,
                            -- references bindings.oid.
  sig_digest text,          -- The digest of the signed message.
  origin text,              -- String describing who initially fed
                            -- the signature to gpg (e.g. "email:claws").
  sig_time integer,         -- Timestamp from the signature.
  time integer,             -- Time this record was created.
  primary key (binding, sig_digest, origin)

GNU extensions to the S2K algorithm

1 octet - S2K Usage: either 254 or 255. 1 octet - S2K Cipher Algo: 0 1 octet - S2K Specifier: 101 3 octets - “GNU” 1 octet - GNU S2K Extension Number.

If such a GNU extension is used neither an IV nor any kind of checksum is used. The defined GNU S2K Extension Numbers are:

Do not store the secret part at all. No specific data follows.
A stub to access smartcards. This data follows:
  • One octet with the length of the following serial number.
  • The serial number. Regardless of what the length octet indicates no more than 16 octets are stored.

Note that gpg stores the GNU S2K Extension Number internally as an S2K Specifier with an offset of 1000.

Format of the OpenPGP TRUST packet

According to RFC4880 (5.10), the trust packet (aka ring trust) is only used within keyrings and contains data that records the user’s specifications of which key holds trusted introducers. The RFC also states that the format of this packet is implementation defined and SHOULD NOT be emitted to output streams or should be ignored on import. GnuPG uses this packet in several additional ways:

1 octet
Trust-Value (only used by Subtype SIG)
1 octet
Signature-Cache (only used by Subtype SIG; value must be less than 128)
3 octets
Fixed value: “gpg”
1 octet
Signature cache (SIG)
Key source on the primary key (KEY)
Key source on a user id (UID)
1 octet
Key Source; i.e. the origin of the key:
Unknown source.
Public keyserver.
Preferred keyserver.
Web Key Directory.
Import from a trusted URL.
Import from a trusted file.
Self generated.
4 octets
Time of last update. This is a a four-octet scalar with the seconds since Epoch.
1 octet
Scalar with the length of the following field.
N octets
String with the URL of the source. This may be a zero-length string.

If the packets contains only two octets a Subtype of 0 is assumed; this is the only format recognized by GnuPG versions < 2.1.18. Trust-Value and Signature-Cache must be zero for all subtypes other than SIG.

Keyserver helper message format

This information is obsolete (Keyserver helpers have been replaced by dirmngr)

The keyserver may be contacted by a Unix Domain socket or via TCP.

The format of a request is:

"Content-length:" digits

Where command-tag is

GET <user-name>
DELETE <user-name>

The format of a response is:

"GNUPG/1.0" status-code status-text
"Content-length:" digits

followed by <digits> bytes of data

Status codes are:

Informational - Request received, continuing process
Success - The action was successfully received, understood, and accepted
Client Error - The request contains bad syntax or cannot be fulfilled
Server Error - The server failed to fulfill an apparently valid request

Object identifiers

OIDs below the GnuPG arc:          GnuPG          notation          pkaAddress          X.509 extensions          standaloneCertificate          wellKnownPrivateKey   invalid encoded OID

Debug flags

This tables gives the flag values for the –debug option along with the alternative names used by the components.


Description of some debug flags:

Used by scdaemon to trace the APDUs exchange with the card.
Show execution times of certain functions.
Trace crypto operations.
Create files with the hashed data.
Trace the Assuan commands.
Show the values of the MPIs.
Used by scdaemon to trace card reader related code. For example: Open and close reader.

Miscellaneous notes

v3 fingerprints

For packet version 3 we calculate the keyids this way:

Low 64 bits of n
Build a v3 pubkey packet (with CTB 0x99) and calculate a RMD160 hash value from it. This is used as the fingerprint and the low 64 bits are the keyid.

Simplified revocation certificates

Revocation certificates consist only of the signature packet; “–import” knows how to handle this. The rationale behind it is to keep them small.

Documentation on HKP (the http keyserver protocol):

A minimalistic HTTP server on port 11371 recognizes a GET for /pks/lookup. The standard http URL encoded query parameters are this (always key=value):

  • op=index (like pgp -kv), op=vindex (like pgp -kvv) and op=get (like pgp -kxa)
  • search=<stringlist>. This is a list of words that must occur in the key. The words are delimited with space, points, @ and so on. The delimiters are not searched for and the order of the words doesn’t matter (but see next option).
  • exact=on. This switch tells the hkp server to only report exact matching keys back. In this case the order and the “delimiters” are important.
  • fingerprint=on. Also reports the fingerprints when used with ‘index’ or ‘vindex’

The keyserver also recognizes http-POSTs to /pks/add. Use this to upload keys.

A better way to do this would be a request like:


This can be implemented using Hurd’s translator mechanism. However, I think the whole keyserver stuff has to be re-thought; I have some ideas and probably create a white paper.

Algorithm names for the “keygen.algo” prompt

When using a –command-fd controlled key generation or “addkey” there is way to know the number to enter on the “keygen.algo” prompt. The displayed numbers are for human reception and may change with releases. To provide a stable way to enter a desired algorithm choice the prompt also accepts predefined names for the algorithms, which will not change.

rsa+rsa1RSA and RSA (default)
dsa+elg2DSA and Elgamal
dsa3DSA (sign only)
rsa/s4RSA (sign only)
elg5Elgamal (encrypt only)
rsa/e6RSA (encrypt only)
dsa/*7DSA (set your own capabilities)
rsa/*8RSA (set your own capabilities)
ecc+ecc9ECC and ECC
ecc/s10ECC (sign only)
ecc/*11ECC (set your own capabilities)
ecc/e12ECC (encrypt only)
keygrip13Existing key
cardkey14Existing key from card

If one of the “foo/*” names are used a “keygen.flags” prompt needs to be answered as well. Instead of toggling the predefined flags, it is also possible to set them direct: Use a “=” character directly followed by a combination of “a” (for authentication), “s” (for signing), or “c” (for certification).

You can’t perform that action at this time.