-
Notifications
You must be signed in to change notification settings - Fork 9
Expand file tree
/
Copy pathgithub_issues.jsonl
More file actions
13 lines (13 loc) · 10.3 KB
/
Copy pathgithub_issues.jsonl
File metadata and controls
13 lines (13 loc) · 10.3 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
{"external_id":"PRM-001","title":"[PRM-001] Freeze the Codex-native architecture boundary","milestone":"M0 Codex-native MVP","labels":["type:architecture","area:skill","priority:P0"],"depends_on":[],"body_markdown":"## Goal\nRecord that Codex is the only coding and execution harness.\n\n## Acceptance criteria\n- [x] ADR selects Skill-first, Plugin-distributed architecture\n- [x] Coding-agent, worktree, sandbox, and GitHub harnesses are outside the published package\n- [x] Evidence-integrity limitations are explicit\n\n## Deliverables\n- docs/adr/0003-codex-native.md\n- docs/architecture.md"}
{"external_id":"PRM-002","title":"[PRM-002] Create the focused PRman Skill","milestone":"M0 Codex-native MVP","labels":["type:engineering","area:skill","priority:P0"],"depends_on":["PRM-001"],"body_markdown":"## Goal\nDefine the reusable Codex workflow without repeating generic coding guidance.\n\n## Acceptance criteria\n- [x] Skill routes assessment, scorer, and write-safety details progressively\n- [x] Review-only requests do not authorize edits\n- [x] GitHub writes require exact human confirmation and remain Draft-only\n- [x] Representative fresh-task behavior is forward-tested in Codex\n\n## Deliverables\n- skills/prman/\n- docs/plugin-installation-validation-2026-08-30.md\n- docs/confirmation-path-validation-2026-08-30.md"}
{"external_id":"PRM-003","title":"[PRM-003] Ship deterministic assessment contracts and CLI","milestone":"M0 Codex-native MVP","labels":["type:engineering","area:core","priority:P0"],"depends_on":["PRM-001"],"body_markdown":"## Goal\nConsume Codex-supplied evidence and return a strict ready/revise/abstain result.\n\n## Acceptance criteria\n- [x] Duplicate and unknown contract fields fail closed\n- [x] Missing, unknown, and fatal gates dominate scorer output\n- [x] Single and comparison modes are deterministic\n- [x] Result never authorizes an external write\n\n## Deliverables\n- src/prman/assessment.py\n- src/prman/decision.py\n- schemas/"}
{"external_id":"PRM-004","title":"[PRM-004] Package PRman as a validated Codex Plugin","milestone":"M0 Codex-native MVP","labels":["type:engineering","area:plugin","priority:P0"],"depends_on":["PRM-002","PRM-003"],"body_markdown":"## Goal\nMake the Skill and deterministic helper installable as one plugin bundle.\n\n## Acceptance criteria\n- [x] Manifest contains valid metadata and the Skill path\n- [x] Bundled helper resolves paths relative to the installed plugin\n- [x] Internal distribution smoke tests pass\n- [x] Official Plugin and Skill validators pass\n\n## Deliverables\n- .codex-plugin/plugin.json\n- skills/prman/scripts/assess.py"}
{"external_id":"PRM-005","title":"[PRM-005] Publish scorer provider conformance tests","milestone":"M1 Production scorer boundary","labels":["type:engineering","area:scorer","priority:P0"],"depends_on":["PRM-003"],"body_markdown":"## Goal\nLet independently packaged scorers prove protocol compatibility.\n\n## Acceptance criteria\n- [ ] Extract reusable provider conformance tests for metadata stability, digest echo, canonical criteria, finite ranges, OOD, authentication, and error handling\n- [x] Test providers are core-classified and always force the final selection to abstain\n- [x] Provider failures produce a structured fail-closed handoff\n\n## Deliverables\n- scorer conformance test module\n- provider author documentation"}
{"external_id":"PRM-006","title":"[PRM-006] Integrate one external calibrated scorer","milestone":"M1 Production scorer boundary","labels":["type:integration","area:scorer","priority:P0"],"depends_on":["PRM-005"],"body_markdown":"## Goal\nExercise PRman with a separately owned, pinned production scorer.\n\n## Acceptance criteria\n- [ ] Provider package or loopback service is pinned by exact revision\n- [ ] Calibrator revision and uncertainty behavior are documented\n- [ ] No model weights or inference dependencies enter the PRman core package\n- [ ] No dataset or model download occurs implicitly\n\n## Deliverables\n- production scorer configuration example\n- redacted integration evidence"}
{"external_id":"PRM-007","title":"[PRM-007] Evaluate evidence integrity and abstention behavior","milestone":"M1 Production scorer boundary","labels":["type:research","area:security","priority:P1"],"depends_on":["PRM-005"],"body_markdown":"## Goal\nMeasure failures caused by missing, stale, contradictory, or fabricated gate evidence.\n\n## Acceptance criteria\n- [ ] Adversarial fixtures cover skipped tests, stale diffs, fatal plus recoverable failures, and truncated context\n- [ ] Abstention rate and false-ready cases are reported\n- [ ] Findings update the Skill and threat model without adding a command harness\n\n## Deliverables\n- evidence-integrity evaluation report\n- regression fixtures"}
{"external_id":"PRM-008","title":"[PRM-008] Test local Plugin installation in fresh Codex tasks","milestone":"M2 Plugin release","labels":["type:testing","area:plugin","priority:P0"],"depends_on":["PRM-004"],"body_markdown":"## Goal\nVerify discovery, invocation, path resolution, and behavior outside the authoring task.\n\n## Acceptance criteria\n- [x] Plugin installs from a local marketplace source\n- [x] Explicit $prman invocation works in a fresh task\n- [x] Implicit invocation triggers for code-change readiness but not generic code questions\n- [x] Uninstall and reinstall pick up a cache-busted update\n\n## Deliverables\n- [x] Redacted installation transcript: docs/plugin-installation-validation-2026-08-30.md\n- [x] Representative task results recorded in the same validation document"}
{"external_id":"PRM-009","title":"[PRM-009] Validate the Draft PR confirmation handoff","milestone":"M2 Plugin release","labels":["type:security","area:github","priority:P0"],"depends_on":["PRM-008"],"body_markdown":"## Goal\nProve that readiness and GitHub write authority remain separate in Codex.\n\n## Acceptance criteria\n- [ ] No write occurs without confirmation naming exact OWNER/REPO and operation\n- [ ] Only a Draft PR is proposed\n- [ ] Merge, auto-merge, force push, and default-branch write are never attempted\n- [ ] A denied confirmation ends with a local handoff\n\n## Deliverables\n- confirmation-path test record\n- updated safety reference"}
{"external_id":"PRM-010","title":"[PRM-010] Record publication and license decisions","milestone":"M2 Plugin release","labels":["type:governance","area:release","priority:P0"],"depends_on":["PRM-008","PRM-009"],"body_markdown":"## Goal\nKeep the already selected public repository and license recorded while completing the release checks.\n\n## Acceptance criteria\n- [x] Canonical public repository is primorLee/PRman\n- [x] Apache-2.0 license is selected and added\n- [x] Historical local prototype files are excluded from the distribution and tracked release surface\n- [ ] Plugin publication receives explicit approval after PRM-008 and PRM-009 complete\n\n## Deliverables\n- LICENSE\n- release checklist\n- approved publication record"}
{"external_id":"PRM-011","title":"[PRM-011] Expand PRman into the end-to-end Draft PR workflow","milestone":"M0 Codex-native MVP","labels":["type:engineering","area:skill","area:github","priority:P0"],"depends_on":["PRM-002","PRM-003","PRM-004"],"body_markdown":"## Goal\nCoordinate discovery, implementation, assessment, confirmation, Draft PR publication, and CI follow-up through Codex.\n\n## Acceptance criteria\n- [x] Skill discovers and selects one contribution-friendly repository or issue read-only\n- [x] Repository instructions, contribution rules, security policy, existing work, base commit, and CI are inspected\n- [x] Codex implements and verifies locally before PRman's deterministic assessment\n- [x] Exact confirmation-packet schema separates assessment from GitHub authority\n- [x] Plugin declares the GitHub MCP dependency\n- [x] Publication is Draft-only and CI repairs are bounded, reassessed, and reconfirmed on material scope change\n- [ ] A clean Codex task validates the complete workflow against a controlled GitHub repository\n\n## Deliverables\n- skills/prman/SKILL.md\n- skills/prman/references/orchestration.md\n- skills/prman/references/github-workflow.md\n- schemas/confirmation_packet.schema.json"}
{"external_id":"PRM-012","title":"[PRM-012] Enforce confirmation and CI transitions with executable contracts","milestone":"M0 Codex-native MVP","labels":["type:engineering","area:core","area:github","priority:P0"],"depends_on":["PRM-003","PRM-011"],"body_markdown":"## Goal\nTurn the prose-only publication boundary into deterministic local acceptance and sequencing checks without adding a GitHub client.\n\n## Acceptance criteria\n- [x] Exact packet and patch digests bind the target, route, diff, PR plan, writes, and CI budget\n- [x] Only the exact target-specific user phrase creates a scoped write authorization\n- [x] Serialized authorization cross-fields are revalidated fail-closed\n- [x] Draft PR URL, number, current CI commit, repair count, new diff, new commit, and in-scope state are checked\n- [x] JSON Schemas, CLI commands, installed-Skill wrapper, and regression tests cover the state flow\n- [x] Identity, remote-observation, and semantic-scope limits are documented\n\n## Deliverables\n- src/prman/workflow.py\n- skills/prman/scripts/workflow.py\n- schemas/confirmation_check.schema.json\n- schemas/write_authorization.schema.json\n- schemas/workflow_run.schema.json\n- tests/core/test_workflow.py"}
{"external_id":"PRM-013","title":"[PRM-013] Keep contribution sessions moving with Codex Goal mode","milestone":"M0 Codex-native MVP","labels":["type:engineering","area:skill","priority:P0"],"depends_on":["PRM-011"],"body_markdown":"## Goal\nAutomatically create or reuse a Codex Goal after intake so PRman continues safe in-scope work without repeatedly asking whether to proceed.\n\n## Acceptance criteria\n- [x] Completed intake becomes one measurable Goal with outcome, constraints, verification, and completion criteria\n- [x] Matching Goals are reused, unrelated Goals are preserved, and no token budget is inferred\n- [x] Search, implementation, testing, review, assessment, CI follow-up, and later contributions continue without needless questions\n- [x] Goal mode does not expand permissions or replace the exact per-PR Draft confirmation\n- [ ] A fresh installed-Plugin task validates automatic Goal creation, confirmation waiting, and resume behavior\n\n## Deliverables\n- skills/prman/references/goal-mode.md\n- skills/prman/SKILL.md\n- docs/architecture.md\n- docs/threat-model.md"}