Skip to content

docs(invariant-checklist): repoint row 10's stale AGENTS.md section citation #174

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
coreytshaffer merged 1 commit into from
Aug 12, 2026
Merged

docs(invariant-checklist): repoint row 10's stale AGENTS.md section citation #174

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion docs/operations/control-plane-invariant-checklist.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -25,7 +25,7 @@ nothing, and is not a certification.
| 7 | Read-only by default; explicit mutation only; exports have no default write location, fail closed on existing files, and repeat byte-identically. | Runtime strategy export path (CR-109/CR-112); workspace write commands require explicit intent | `pytest tests/test_reports.py -q` plus runtime-strategy tests; attempt an export without `--output` | Verified 2026-07-07 (in full suite) |
| 8 | Deterministic fixture reports never ingest recorded or probe data; recorded/probe tiers are always labeled and validated via strict mapping (unknown fields rejected). | `triage_core/runtime_strategy_evidence.py` (CR-104 → CR-112); boundary for future telemetry defined in [local-backend-telemetry.md](local-backend-telemetry.md) | Runtime-strategy focused tests; confirm `recorded-report` requires explicit `--input` | Verified 2026-07-07 (in full suite) |
| 9 | Failure reporting uses closed vocabularies (reason codes), never raw error text; quality gates qualify cost interpretations but never rewrite or rank them. | CR-106/CR-108 vocabularies; CR-097 categories; CR-113 brief for future telemetry | Grep CLI output in tests for reason codes; confirm no "best strategy" output exists | Verified 2026-07-07 (in full suite) |
| 10 | Human review gates stay outside every automated loop; no evaluator, supervisor, or agent verdict becomes approval authority. | Workflow doctrine (README Safety Invariants, AGENTS.md §7); `tc doctor` runtime-safety postures | `tc doctor` → external execution `blocked`, human approval `human-review-required`, network/tool execution `unavailable` | Verified 2026-07-07 (doctor output) |
| 10 | Human review gates stay outside every automated loop; no evaluator, supervisor, or agent verdict becomes approval authority. | Workflow doctrine (README Safety Invariants, AGENTS.md, "Agent verdicts are not authority"); `tc doctor` runtime-safety postures | `tc doctor` → external execution `blocked`, human approval `human-review-required`, network/tool execution `unavailable` | Verified 2026-07-07 (doctor output) |

## Rules for Future Slices

Expand Down
Loading