Repository navigation
Releases
Releases · tekmora/epitrop
v0.2.0
This release ships the code of v0.2.0-rc.1 unchanged.
Added
- A sandboxed job can reach a list of declared destinations and nothing else:
sandbox.network: proxywithsandbox.destinations(#25) - Add signed unit checkpoints for fast replay, preemption at slot safe points, and refusal of diverged log copies (#11)
- BREAKING: Interval and event job triggers in supervisor, refusing unexecutable schedules or restart policies on triggered jobs at config load (#126)
- Take up interrupted slot runs with epitropctl slot resume, adopting the running or ended job without a second start and recording the result as an uninterrupted run (#127)
- Traces and metrics for slot runs, supervised jobs, side effects and task-graph queue depth, exported over OTLP when an endpoint is set (#34)
- Start a slot run as a systemd user service of its own with epitropctl slot run --detach, carrying the template's memory_max and memory_high, so the run outlives the session that started it
- units frontier, assign and release: list the graph nodes a slot may take, claim one and open its unit in one record, and give it back with a reason
- units handover: the open units of one repository with their pending decisions and unread notes, written as a slot's resume note on request
epitropctl watchaccepts--github-base,--huggingface-base, and--ntfy-baseflags to point API requests at local stand-ins (#145)- A sandboxed job can declare paths inside its workspace read-only, and cannot rename them away (#128)
- End a hung job whose heartbeat, the last write to its log, is older than heartbeat.max_age_seconds, recording job.stale and leaving the restart to its policy (#21)
- Sync reads the calling LiteLLM key's budget window, and epitropctl headroom reports a stale or missing usage reading as unknown, never as zero (#24)
- Cutover 1 plan: the fleet's background jobs declared as Epitrop jobs in
docs/cutover-1.example.yaml, with the parallel run and switch-over criteria indocs/cutover-1.md(#36) docs/config.example.yaml: running the release watch's supervisor as a systemd user unit that loads its token withLoadCredentialEncrypted=, with the token's minimal rights (#23)- Release watch: every pass ends with one summary line (sources polled and failed; releases new, filed and deferred; whether the rate-limit throttle stopped it), so a pass without new releases still shows in the job log (#271)
- Catalog: the usage window carries
burn_percent_per_hourandactive_agents(optional, at least zero; schema and validation refuse anything else); codex-local measures both over the last 15 minutes of session logs, and litellm-key-budget names both as absent. The schema stays at version 1, so a reader that validates against the v0.1.0 schema refuses snapshots carrying the new fields until it takes the new schema (#274) - BREAKING: Sync: the claude-local source reads Claude Code session logs and writes the five_hour and seven_day usage windows; used_percent and the burn rate are derived from a ceiling observed at a rejected request and stay absent with a reason until one is seen, and epitropctl claude-limits lists the limit types the logs name; a snapshot may now hold 25,000 records, which a reader from an earlier release refuses (#276)
Changed
- BREAKING: Loading a configuration refuses a job or watch whose
router_aliasis not listed inrouter.aliases - BREAKING: Allow a sandboxed job's workspace and inputs only inside the directories sandbox_input_roots declares, and refuse login locations below HOME unless a root names exactly one
- Cutover 1 parallel run: shadow jobs for the landing daemon, auto-rebase, the update intake and the monitor run each script's dry mode with a heartbeat bound (#36)
docs/cutover-1.md: the shadow supervisor's host unit tags its journal lines with its own name, so its exit reason can be found byjournalctl -t(#263)- BREAKING: The /events HTTP stream serves signed event-log records in sequence order, replacing unit and job diff summaries; each event carries the record sequence number as its id and the event type as its event name, and clients can resume from any record with Last-Event-ID (#278)
Fixed
- A sandboxed job gets its arguments as its config wrote them:
systemd-runno longer replaces${NAME}and$$in them, and the value of an allowed variable no longer lands on the argv (#25) - A run-record, task-graph or effect-resolution write that loses a race to another writer is now refused instead of reported as written (#108)
- A sandboxed job with egress network no longer logs pasta's "Couldn't get any nameserver address" line (#137)
- A release watch run without GITHUB_TOKEN now names the missing token when a sink search is refused, and logs once when rate_limit_floor is at or above the GitHub rate limit (#138)
epitropctl watchrefuses a release watch seen-set file whose version is not 1, instead of reading it and rewriting it as version 1 (#148)- The signed event log and the epitropctl commands that write it work on Windows, and Windows test suites run every package (#152)
- Stopping a job whose shim is exiting no longer fails with "job lock is not held"; the job counts as stopped (#227)
- A sandboxed job start is no longer refused when the kernel still shows the shim's program for the job right after the start. (#228)
epitropctl jobs superviseno longer ends after 30 seconds with a deadline error; it runs until SIGTERM or an interrupt and then exits 0 (#36)
Migration
- Snapshots (#276, #277): upgrade every program that reads snapshots before a writer selects sources whose caps sum past 20,000 records. Code that only compares against
catalog.MaxSnapshotRecordsneeds no change. - Event stream (#278, #279):
/eventsno longer sends the unit and job summary events. Read event names as record types (unit.recorded,job.started, ...), parsedataas the record, and read the current state back with the tools after a 410.
v0.2.0-rc.1
Added
- A sandboxed job can reach a list of declared destinations and nothing else:
sandbox.network: proxywithsandbox.destinations(#25) - Add signed unit checkpoints for fast replay, preemption at slot safe points, and refusal of diverged log copies (#11)
- BREAKING: Interval and event job triggers in supervisor, refusing unexecutable schedules or restart policies on triggered jobs at config load (#126)
- Take up interrupted slot runs with epitropctl slot resume, adopting the running or ended job without a second start and recording the result as an uninterrupted run (#127)
- Traces and metrics for slot runs, supervised jobs, side effects and task-graph queue depth, exported over OTLP when an endpoint is set (#34)
- Start a slot run as a systemd user service of its own with epitropctl slot run --detach, carrying the template's memory_max and memory_high, so the run outlives the session that started it
- units frontier, assign and release: list the graph nodes a slot may take, claim one and open its unit in one record, and give it back with a reason
- units handover: the open units of one repository with their pending decisions and unread notes, written as a slot's resume note on request
epitropctl watchaccepts--github-base,--huggingface-base, and--ntfy-baseflags to point API requests at local stand-ins (#145)- A sandboxed job can declare paths inside its workspace read-only, and cannot rename them away (#128)
- End a hung job whose heartbeat, the last write to its log, is older than heartbeat.max_age_seconds, recording job.stale and leaving the restart to its policy (#21)
- Sync reads the calling LiteLLM key's budget window, and epitropctl headroom reports a stale or missing usage reading as unknown, never as zero (#24)
- Cutover 1 plan: the fleet's background jobs declared as Epitrop jobs in
docs/cutover-1.example.yaml, with the parallel run and switch-over criteria indocs/cutover-1.md(#36) docs/config.example.yaml: running the release watch's supervisor as a systemd user unit that loads its token withLoadCredentialEncrypted=, with the token's minimal rights (#23)- Release watch: every pass ends with one summary line (sources polled and failed; releases new, filed and deferred; whether the rate-limit throttle stopped it), so a pass without new releases still shows in the job log (#271)
- Catalog: the usage window carries
burn_percent_per_hourandactive_agents(optional, at least zero; schema and validation refuse anything else); codex-local measures both over the last 15 minutes of session logs, and litellm-key-budget names both as absent. The schema stays at version 1, so a reader that validates against the v0.1.0 schema refuses snapshots carrying the new fields until it takes the new schema (#274) - BREAKING: Sync: the claude-local source reads Claude Code session logs and writes the five_hour and seven_day usage windows; used_percent and the burn rate are derived from a ceiling observed at a rejected request and stay absent with a reason until one is seen, and epitropctl claude-limits lists the limit types the logs name; a snapshot may now hold 25,000 records, which a reader from an earlier release refuses (#276)
Changed
- BREAKING: Loading a configuration refuses a job or watch whose
router_aliasis not listed inrouter.aliases - BREAKING: Allow a sandboxed job's workspace and inputs only inside the directories sandbox_input_roots declares, and refuse login locations below HOME unless a root names exactly one
- Cutover 1 parallel run: shadow jobs for the landing daemon, auto-rebase, the update intake and the monitor run each script's dry mode with a heartbeat bound (#36)
docs/cutover-1.md: the shadow supervisor's host unit tags its journal lines with its own name, so its exit reason can be found byjournalctl -t(#263)- BREAKING: The /events HTTP stream serves signed event-log records in sequence order, replacing unit and job diff summaries; each event carries the record sequence number as its id and the event type as its event name, and clients can resume from any record with Last-Event-ID (#278)
Fixed
- A sandboxed job gets its arguments as its config wrote them:
systemd-runno longer replaces${NAME}and$$in them, and the value of an allowed variable no longer lands on the argv (#25) - A run-record, task-graph or effect-resolution write that loses a race to another writer is now refused instead of reported as written (#108)
- A sandboxed job with egress network no longer logs pasta's "Couldn't get any nameserver address" line (#137)
- A release watch run without GITHUB_TOKEN now names the missing token when a sink search is refused, and logs once when rate_limit_floor is at or above the GitHub rate limit (#138)
epitropctl watchrefuses a release watch seen-set file whose version is not 1, instead of reading it and rewriting it as version 1 (#148)- The signed event log and the epitropctl commands that write it work on Windows, and Windows test suites run every package (#152)
- Stopping a job whose shim is exiting no longer fails with "job lock is not held"; the job counts as stopped (#227)
- A sandboxed job start is no longer refused when the kernel still shows the shim's program for the job right after the start. (#228)
epitropctl jobs superviseno longer ends after 30 seconds with a deadline error; it runs until SIGTERM or an interrupt and then exits 0 (#36)
Migration
- Snapshots (#276, #277): upgrade every program that reads snapshots before a writer selects sources whose caps sum past 20,000 records. Code that only compares against
catalog.MaxSnapshotRecordsneeds no change. - Event stream (#278, #279):
/eventsno longer sends the unit and job summary events. Read event names as record types (unit.recorded,job.started, ...), parsedataas the record, and read the current state back with the tools after a 410.
v0.1.0
Changed
- No change since 0.1.0-rc.3: the first release is the code of that candidate, and the changelog sections of the release candidates list what it holds. Known gap: a sandboxed job with the network mode egress reaches the whole internet, and no mode limits a job to declared destinations. A job started by slot run that needs its model gateway therefore has network access to every address. Run no real work unit in a slot until a later release closes this (#25)
v0.1.0-rc.3
v0.1.0-rc.2
Changelog
- aab1b8b chore(deps): revert berriai/litellm digest to 7ff5010 (#101)
- 1756977 chore(deps): update berriai/litellm digest to 18e87d3 (#68)
- 34cf128 chore(deps): update berriai/litellm digest to 7ff5010 (#74)
- ae95072 chore(deps): update berriai/litellm digest to 988cc7d (#88)
- 5a4cd72 chore(deps): update berriai/litellm digest to a3236ba (#50)
- 2fd6e33 chore(deps): update berriai/litellm digest to fff86d6 (#71)
- 2a2597a chore(deps): update golang:1.27-trixie docker digest to 2f84bc9 (#43)
- 919f223 chore(deps): update ollama/ollama digest to 021cda5 (#69)
- 86afa7b chore(deps): update ollama/ollama digest to eab97e9 (#72)
- 575cd7e chore(deps): update openai/codex digest to 2351d9e (#51)
- 7d834f9 chore(deps): update openai/codex digest to 52950c5 (#73)
- 44c879c chore(deps): update openai/codex digest to 806d973 (#76)
- 6298b0d chore(deps): update openai/codex digest to 8ec80de (#70)
- b4df405 chore(deps): update openai/openai-openapi digest to e95c0fe (#75)
- 49e2931 chore(release): render the changelog for v0.1.0-rc.2 (#117)
- 3a4954f chore(renovate): wait three days before adopting a release (#105)
- f8869a6 chore: add changelog fragments for the changes since rc.1 (#102)
- f54680e chore: adopt Praetor standards (#41)
- 9f19009 chore: configure Renovate (#18)
- 937eb9a chore: declare Praetor-managed files to Renovate (#45)
- 826da5e chore: make the catalog slice a standalone Go module
- f6f1d3d chore: re-adopt Praetor at 9615f1b and pin the engine (#81)
- 232515e chore: re-adopt Praetor at a259c51 (#67)
- 0fc1d45 chore: re-adopt Praetor at e9b102a (#78)
- d6213f6 chore: re-adopt Praetor to build the API checker from a pinned module (#63)
- 9b0be3a chore: release v0.1.0-rc.1 (#65)
- e721ec2 ci: build a release snapshot on every change that can break it (#103)
- 3c42285 ci: check that go.sum is tidy, and keep update PRs from untidying it (#96)
- 59f1128 ci: require govulncheck and gosec on every pull request (#57)
- 115312d ci: run the adopted gates on every pull request (#52)
- ad8e0a3 feat!: rename the runtime to Epitrop under tekmora/epitrop (#110)
- 8a2f2e8 feat(tribunus): add tribunusctl sync/show CLI, data sync slice 1 (#301)
- 21f2f28 feat: add the RT-01 check for a model call, with fixtures both ways (#113)
- 5bdd770 feat: add the acyclic task graph with staged, bounded changes (#66)
- d2d2360 feat: add the workspace and the result contract of an agent slot (#115)
- 1438451 feat: append-only event log with signed records and replay (#62)
- e5da4d4 feat: base the control plane on golusoris, pinned (#47)
- ec8d7af feat: check catalog sources against pinned upstream schemas (#49)
- 28037cc feat: give a sandboxed job an optional soft memory limit (#98)
- 298c614 feat: keep a run record of work units and answer what to resume (#82)
- e043375 feat: keep an evidence pointer on the unit record (#99)
- 71fcd75 feat: load config from one schema-checked file (#59)
- 0faaed8 feat: publish signed release artefacts for tribunusctl (#55)
- a293fed feat: rebuild the task graph from the event log (#107)
- 4cdec6c feat: record a refused job start as an event (#94)
- 6a8f8a9 feat: run each outbound side effect at most once, from a ledger in the event log (#112)
- 1d43aec feat: run one work unit in a slot with epitropctl slot run (#116)
- b0411ec feat: run supervised jobs in a sandbox (#79)
- e6da1a2 feat: stop a job at its own deadline (#95)
- a2d374e feat: supervise jobs with a per-job shim and a validated stop path (#77)
- a7e698a feat: version the catalog snapshot schema (#54)
- 779cf03 feat: watch upstream releases and file one issue per release (#80)
- 4157dff feat: write the task graph to the event log, with graph show and graph apply (#109)
- c11daec fix(deps): update module github.com/golusoris/golusoris to v0.16.2 (#85)
- b4641c9 fix(deps): update module github.com/golusoris/golusoris/core to v0.10.2 (#84)
- afcac86 fix(deps): update module golang.org/x/sys to v0.49.0 (#86)
- 0591fe4 fix(tribunus): frame data sync as graph-router input (#384)
- 24e9190 fix: apply the configured event log lock timeout in the supervisor and the shim (#106)
- ce0485a fix: bound the supervisor tests by the code's own waits and end what they start (#104)
- 41ca0d3 fix: collect the shims a supervisor starts (#92)
- 6e72512 fix: land 2 reviewed batches (close Platform Neutrality gaps and the H; neutralize litellm gateway defaults, fix) (#469)
- 70006ef fix: let schemas-repin move a pinned upstream document (#56)
- 558a270 fix: let the supervisor tests compile off Linux (#97)
- 63b95d5 fix: make the start-wait tests independent of the machine's process ids (#91)
- e8df7ff fix: no silent fallback in the catalog sources and sync (#48)
- 7628754 fix: publish release-candidate tags as prereleases (#64)
- 92dc653 fix: refuse a sandboxed job found outside its systemd scope (#87)