Repository navigation
v0.3.0-alpha.19
Pre-releaseAdded
- The Python SDK gains
judge(task_id, judgement_id, verdict, *, identity, rationale=None), which answers a waiting task's semantic question throughmelra_task_judgeas the TypeScript SDK'sjudge()does (#378). - A settled semantic certificate's answered item carries
executedBy, the action receipt's declared principal, besidejudgedBy, so the two declared chains can be compared without the receipt (#378). Both are claims, and neither marks theharness:link an OAuth client's token adds. It is redacted with the certificate; certificates written before it simply lack it, and nothing is migrated.
Changed
-
Breaking:
melra_task_judgealso refuses a judge whose declared delegation chain names the task's principal, such as a grader the acting agent started (semantic_judge_is_actor, #378). Another session under the same harness, or another agent of the person the work was for, may still answer, unless the task's principal is that harness or person itself. Theharness:link an OAuth client's token puts above a judge's chain is recorded injudgedBybut not compared. The check still rests on declared identities;melra judgeis unchanged. -
Breaking:
melra approvewith an empty task id, as an unset"$ID"sends, is refused withtask_not_found:instead of taking the one task awaiting approval, if only one was; a prefix no waiting task has is refused withtask_not_found:<prefix>(#386). -
A task's goal is stored as declared, with only recognisable secret shapes taken out, instead of being replaced whole (#385). The stored request, the certificate's
goal, and a workflow definition's step and delegation goals used to read[REDACTED_SENSITIVE_FIELD]whatever the goal said; they now hold the goal itself, so a certificate states what its task was for. The trade-off: a secret with no recognisable shape written into a goal is now kept in those records, so keep secrets out of goals; the goal's schema description now says so to whoever writes one. Thebrowseharness tool's goal no longer includes the URL's userinfo, query or fragment. Tasks planned before this keep the placeholder, as do certificates later written for them. A definition stored before this keeps it in its stored copy (bump its version to store the goal there), but a new run of it reads the sealed definition, so that run's tasks, certificates and delegation prompts hold the goal. Nothing is migrated. -
Breaking: an approval challenge's
actionDigest(and so its phrase, stillAPPROVEplus its first twelve hex digits), therenderDigestan approval is bound to, and a receipt'sinputDigest, which now also binds the task id, can no longer be recomputed without the home's payload key. A file read's or write'ssha256and a failedfile_hashitem'sdigestare stored as[REDACTED_SENSITIVE_FIELD]. The caller ofmelra_executestill receives thesha256inoutput, but a failedfile_hashitem'sdigestis masked in the receipt and certificate and is no longer returned anywhere; run a filehashoperation to read the actual hash. In a home shared with0.3.0-alpha.17or0.3.0-alpha.18, this version still runs challenges and accepts decisions those versions recorded, and a decision it records for a challenge one of them issued keeps the plainrenderDigesttheir executor checks; an older process refuses a challenge this version issued withapproval_action_digest_mismatchbefore anything runs, and the task waits for a newer process. Upgrade every process sharing aMELRA_HOMEtogether and re-runmelra setup(compatibility).
Fixed
melra inspectandmelra exporttake the eight-character id prefix the plainmelra taskstable prints, in any status, asmelra approvealready did. Before, they passed it on as a whole id and reportedtask_not_found. A prefix two tasks share is refused withtask_id_ambiguous:<prefix>, and one no task has withtask_not_found:<prefix>; an empty argument names no task.melra judgestill takes ids whole. (#386)examples/README.mdsaid the API example succeeds unauthenticated withoutGITHUB_TOKEN; it fails withcredential_source_missing:githubbefore any request is sent, as it always has. The example now says it needs the token, anddocs/INSTALLATION.md#credentialsdocumentscredential_source_missingbesidecredential_source_invalid: a credential whose source is missing fails every call to its hosts and never falls back to an unauthenticated one. (#387)- The
recallharness tool no longer fails a query it accepts that is longer than 9,993 characters: the goal built from it passed the 10,000 characters a goal may hold. A harness tool's goal that would pass that keeps only its verb, as inRecall (input too long to restate), rather than being cut through its input; the operation still carries the input (#385). - A verdict sent in band where only an operator's counts, from a developer-mode process on a task that ran enforced (
melra judgewithout--mode enforced, ormelra_task_judge) or from an agent on an enforced server (melra_task_judge), is checked for the task's status, the judgement id, an earlier answer and expiry first, so it getstask_not_awaiting_judgement:<status>,judgement_not_found,judgement_already_answeredorjudgement_expiredwhere one applies; only a verdict on a question still waiting getsjudgement_in_band_refused. Before, every one gotjudgement_in_band_refused, withmelra judgeadvice to rerun with--mode enforced, and only the rerun showed the refusal it would meet. Every one of these refusals still records nothing. (#384) melra judgeon a question past itsexpiresAtreportsjudgement_expired(exit 1), as documented. Its own startup settles the expired wait first, so it reportedtask_not_awaiting_judgement:partial(or:recovery_required).melra tasksandmelra statuscount awaiting_usertask as waiting only while one of its open questions has not expired, and count the expired ones apart, with a hint that they settle closed when a runtime next opens the data directory;status --jsonaddsawaitingJudgementandexpiredJudgements. Before, only the storedwaiting_usercount was reported, so an expired wait counted as waiting until something opened the data directory. (#383)MelraClient.connect()in the TypeScript SDK gives the stdio server it spawns this process's whole environment exceptMELRA_UNHINGEDandMELRA_SOCKET, and exceptMELRA_WORKSPACEwhencwdis given, withenvon top: the Python SDK's rule. It passed onlyMELRA_HOME,MELRA_POLICY,MELRA_MODE,MELRA_PAYLOAD_KEYandMELRA_WORKSPACE, plus the MCP SDK'sHOME,PATHand a few more, so a credential a policy reads fromsource.envwas missing and every credentialed call failed (credential_source_missing),MELRA_BROWSER*were dropped without a word, and terminal commands lostLANG,LC_ALL,CIandNO_COLOR. Both SDK READMEs now say what a spawned server inherits. (#382)
Security
- Digests stored beside redacted records are now keyed per home, with a key derived from the payload key, and a content digest stored beside redacted content is masked like it. A workflow step's idempotency key stays plain, so every version sharing a home derives it alike; the threat model lists what else stays plain. Records written before this are not rewritten.
- A receipt's
principal, the declared delegation chain, is redacted like the rest of the record, as the task row,judgedByand a certificate'sexecutedByalready were, so it reads the same asexecutedBy, and the kinds taken out are listed in the receipt'sredactions. Receipts written before this are left as they were. - The
promptahuman_inputordelegationnode records in a run's state, and so in its events and snapshots, is redacted like the rest of the record, as the stored definition's copy already was. Only secret shapes are taken out, so a delegation's prompt still says what to do, and it stays within the 2,000 characters node state allows. A run's state, events and snapshots written before this keep the prompt they had; a node still waiting takes the redacted prompt at the run's next advance, which records one moreworkflow.node_changed.