Cold review returned BLOCKED with 1 Blocker, 4 High, 4 Medium, 1 Low. The Blocker is
this tool committing the defect it exists to catch: nearly every non-zero status was
read as "the test failed", so a SIGTERM (143) produced PROVEN and exit 0. Cancellation
masquerading as evidence.
Fixed by measurement, not assumption. Exit codes on this machine: pass 0, fail 1,
collection-error 2, SIGTERM 143 — and with --continue-on-collection-errors a collection
error becomes 1. So the script now passes that flag and accepts ONLY exit 1 as red,
which closes the ambiguous 2/3 band without needing to parse output, and still supports
the add-a-file case (reverting the module the test imports). INT/TERM handlers now exit
explicitly; without that the trap fell through and the previous status stood.
Proven load-bearing: test_SIGTERM_mid_run_never_reports_PROVEN goes RED when the
exit-code strictness and the signal exit are mutated out.
The MUTATION grade was NOT earned and is now earned. Review was right: the old test
deleted the clause and asserted the MUTANT misbehaved, leaving the named test green,
which is an argument rather than a witnessed red — and docs/guard-inventory.md says
plainly that the column is not a grading curve. The test now re-runs the UNCHANGED
refusal test against the mutant via PROVE_FIX_PATH and requires it to FAIL (exit 1, one
reported failure, named). Loose assertions were tightened after checking WHY the nested
run failed rather than that it failed.
Other findings, all real:
- control and reverted phases now get SEPARATE worktrees and TMPDIRs, so state written
by the control run cannot decide the second one;
- revert is driven by git's own --name-status letters (NUL-delimited) instead of probing
the parent blob and reading error text — `cat-file -e parent:new.py` writes a non-empty
message for a legitimately added file, so the old logic both misclassified that and
could not distinguish a real git failure;
- the CI loop no longer parses selectors with `xargs` (it mangles a parametrised node id
like test_x[can't] into an empty selector, silently dropping a claim) and no longer
enumerates commits fail-open (`for sha in $(git rev-list ...)` succeeds with zero
commits when git fails);
- merge commits carrying `Proves:` are refused with a reason instead of being skipped by
`--no-merges`;
- PROVEN now prints the observed failures, since the docstring asks review to judge
whether the RIGHT property failed and hiding the report made that impossible.
Also corrected an overclaim of my own: the self-proof on the previous commit goes red
because the negative phase DELETES prove-fix.sh, so it demonstrates the file is required,
not that the logic discriminates. The clause mutation is what carries that.
Verified: 13 tests; real commits 84165ab75 PROVEN / UNPROVEN against an unrelated test /
merge commit refused (rc=5); scripts/tests 680 passed, 2 skipped.
Proves: scripts/tests/test_prove_fix.py
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
docs/ — task-signal map
Purpose: route a fresh contributor/agent to the minimal set of docs for the task at hand, instead of a mandatory front-to-back read. Update this doc in the same PR that adds, removes, or retitles a doc below, or that changes which sections a task signal points to.
Start here, always
CLAUDE.md(repo root) — project intro: architecture, layout, dev commands, conventions, Task Completion Protocol.docs/contributing.md— established code patterns (CQRS/MediatR, LanguageExt, the ChicoryTV SPA, EF Core dual-provider migrations, FFmpeg pipeline, analyzers, testing). Read before any non-trivial change.
Task signal → minimal sections
| Signal | Read |
|---|---|
| Session startup / "what's next" (no issue named) | docs/handoffs/chicorytv-issue-queue.md (standing kickoff — two concurrent tracks: orientation ‖ scripts/select-queue.sh 5) |
| Named-issue pickup | Skip queue selection; go straight to focused retrieval — see "Knowledge retrieval" below, then the issue body |
Adding/changing a /api/* endpoint |
docs/api-conventions.md checklist + docs/endpoint-index.md |
| Adding a ChicoryTV SPA screen | docs/spa-conventions.md |
| Scheduling / playout engine work | docs/domain-model.md + decisions catalog rows keyed sched.* (docs/decisions/README.md) |
| Concurrency / optimistic-locking work | docs/api-conventions.md §7a/b/c + docs/decisions/optimistic-concurrency.md |
| Auth / security-surface work | docs/decisions/api-auth-security.md |
| CI / release pipeline work | docs/ci-cd.md + docs/decisions/release-ci-governance.md |
| Proposing a new guard / CI check / regression test convention | docs/defect-shapes-773.md §4 (detector menu + the classes where no detector is plausible), then the two rules every guard must satisfy: docs/decisions/records/testing/guard-derives-population-from-source.md and …/guard-ships-with-mutation-proof.md |
| Adding / changing / deleting a guard file | docs/guard-inventory.md — every guard's row is machine-checked by scripts/tests/test_guard_inventory.py, so a new guard must acquire a row before the suite goes green |
| Finding every site that references a symbol (multi-site fix/sweep) | docs/local-lsp-tooling.md — which surface answers, and why a delegated agent must be pointed at the csharp-lsp MCP tools rather than the LSP tool |
| Live local run / Playwright-MCP verification | docs/e2e-local.md + scripts/e2e-local.sh |
| Adding/changing a UI-E2E browser flow | docs/e2e-local.md → "UI-E2E harness" + scripts/e2e-ui.sh |
| What does a test suite cover | docs/testing.md |
| Legacy Blazor route lookup | docs/blazor-route-parity.md (historical #91 phase (b) inventory) |
| "Why do we do X this way" / challenging a convention | Catalog-first: docs/decisions/README.md (active rows) → follow the row's link to docs/decisions/records/<area>/<topic>.md for full rationale. docs/decisions/archive/<area>/ only for "what did the rule used to be." |
Knowledge retrieval (MemPalace + catalog + Gitea)
These four rules are the seam agreed with server-management#642 (the Gitea→MemPalace exporter). They apply whether the question comes up via MemPalace, a grep, or a stale comment:
- Current conventions/decisions → catalog-first. Start at
docs/decisions/README.md; discover via theErsatzTV-Decisionswing (active) /ErsatzTV-Decisions-Archive(superseded/retired). Resolve by topic/key, never by chasing a file path. - Issue history → evidence, not authority. The
Gitea-ErsatzTVwing is historical narrative that may be stale; it never overrides current Markdown. - The breadcrumb rule (the crux behavior change). A file path named inside a historical issue
comment (e.g. "grep
docs/decisions.md2026-07-17", "see …") is a breadcrumb, not a live pointer. Find the current rule via the catalog / active wing by concept; do not treat the named path as current. (Why it's safe: still-current → in the active wing, breadcrumb resolves; superseded → the active wing returns the successor and a literal follow lands on a record that announces its ownstatus: superseded; retired → the active wing returns nothing, which is itself the signal. The validator-enforced move-to-archive/is what prevents the catastrophic "superseded rule read as current" case.) - Fallback when MemPalace is stale/down:
docs/decisions/README.mdcatalog, thenrg '^`key: <dotted.key>`' docs/decisions/. MemPalace is never authority nor sole fallback.
MemPalace is candidate discovery only — every passage is verified against its cited Markdown/Gitea
source before use. Never derive live queue state from MemPalace, #237, or historical comments; queue
state is live Gitea state, retrieved via scripts/select-queue.sh (see
docs/handoffs/chicorytv-issue-queue.md). Full retrieval contract (altitude/precedence, staleness
bounds, what's mined per issue): docs/handoffs/chicorytv-issue-queue.md → "Knowledge retrieval".
Also present in docs/
docs/domain-model.md— what the app IS: entity glossary, channel→playout→schedule/block concept map, where each concept is edited in the SPA.docs/api-conventions.md— checklist for adding/changing a/api/*endpoint (controllers, DTOs, error mapping, auth, OpenAPI regen, tests).docs/spa-conventions.md— playbook for adding a screen to the ChicoryTV React SPA.docs/e2e-local.md(+scripts/e2e-local.sh) — how to run a live local instance for manual or Playwright-MCP verification.docs/local-lsp-tooling.md— the code-intelligence surfaces (theLSPtool's three servers and thecsharp-lspMCP server): how each is configured, which one a subagent can actually reach, the traps (a cold server answers the first query with a confidently partial result), andscripts/check-local-lsp.shto verify the preconditions. Read before briefing an agent to find every site referencing a symbol.docs/testing.md— testing map: what each*.Testsproject /websuite covers, golden-file nets, the timezone-independence rule, how to run subsets, the per-PR verification gate.docs/blazor-route-parity.md— historical record of the completed #91 phase (b) cutover: the Blazor Server UI is removed and every legacy route now 302-redirects to its SPA equivalent (or falls through to the catch-all →/app). Read it for the full legacy→SPA route inventory.docs/decisions/records/<area>/<topic>.md— one active decision record per file, YAML frontmatter (key/title/status/since/supersedes/superseded-by, plus optionalstale-after/sources— ersatztv#603), rationale prose in the body. The filename is the key, so one-active-record-per-key is a filesystem property (ersatztv#610).docs/decisions.mdand the topic files remain as the lifecycle-schema narrative plus a "Records formerly in this file" index, which is what keeps older date-based pointers resolvable. Generated active view:docs/decisions/README.md(catalog / task router) — start there. Superseded/retired records live indocs/decisions/archive/and are read only for history, never for "what is the current rule."docs/ci-cd.md— build/test/release pipeline, versioning, dependency management.docs/rest-api.md— REST API design doc for ersatztv#2 (goals, conventions, per-slice plan). Largely superseded day-to-day bydocs/api-conventions.md; read this for the original rationale.docs/mcp.md— theErsatzTV.Mcpstdio JSON-RPC MCP server (#58): how it wraps/api/v1as read + cautious-write tools, its config/env vars, auth, security posture, and the tool catalog.docs/channels.md— Channel entity field reference.docs/m3u-xmltv.md— M3U/XMLTV generation overview (ChannelPlaylist,GetChannelGuideHandler).docs/fork-strategy.md— divergence policy vs upstream ErsatzTV.docs/design-sync.md— Claude Design ↔ repo screen workflow (#92).docs/endpoint-index.md— generated REST endpoint index (method/path/operationId/summary per OpenAPI tag). Do not edit by hand; regenerated byscripts/generate-endpoint-index.py/scripts/update-openapi.sh.docs/handoffs/chicorytv-issue-queue.md— static session kickoff prompt + workflow lore. Queue state is live Gitea state, retrieved each session viascripts/select-queue.sh— see that file's standing kickoff for the two concurrent tracks (orientation ‖ selection). ersatztv#237 is a closed, archival historical tracker (superseded bystartup.parallel-orientationindocs/decisions.md) — not a live pointer.docs/defect-shapes-773.md— root-cause analysis of the recurring defect shapes across the whole closed-issue corpus (#773): the measured class ranking, the four families they consolidate into, the cheapest mechanical detector per class, the classes where no detector is plausible, and an audit of which configured hooks/MCP servers/LSPs are actually invoked. Read it before proposing a new guard or CI check — §4 is the detector menu, and it argues against enumerating cases one incident at a time.docs/guard-inventory.md— every executable guard file, what it blocks, whether it is aGUARDorTOOLING, and whether it ships a mutation proof (MUTATION/BEHAVIOUR-ONLY/NONE) with afile::functionref. The population is derived from the filesystem and the workflow/hook call sites and compared for set equality byscripts/tests/test_guard_inventory.py, so a new guard cannot ship unclassified and a renamed test cannot leave a row claiming coverage it has lost. Guards implemented inline in workflow YAML are deliberately outside that population — the doc states the limit rather than implying coverage.docs/tracker-retrofit-triage-237.md— audit trail for the #524 triage of ersatztv#237's 111 comments (method, per-comment classification, totals). Evidence for thedocs.tracker-comment-retrofitdecision; read it only when triaging another over-cap tracker.docs/handoffs/rest-api.md— original handoff prompt for kicking off the REST API work (#2).