feat: add gates ensuring discuss-phase decisions are translated to plans and verified (closes #2492) (#2611)

* feat(#2492): add gates ensuring discuss-phase decisions are translated and verified

Two gates close the loop between CONTEXT.md `<decisions>` and downstream
work, fixing #2492:

- Plan-phase **translation gate** (BLOCKING). After requirements
  coverage, refuses to mark a phase planned when a trackable decision
  is not cited (by id `D-NN` or by 6+-word phrase) in any plan's
  `must_haves`, `truths`, or body. Failure message names each missed
  decision with id, category, text, and remediation paths.

- Verify-phase **validation gate** (NON-BLOCKING). Searches plans,
  SUMMARY.md, files modified, and recent commit subjects for each
  trackable decision. Misses are written to VERIFICATION.md as a
  warning section but do not change verification status. Asymmetry is
  deliberate — fuzzy-match miss should not fail an otherwise green
  phase.

Shared helper `parseDecisions()` lives in `sdk/src/query/decisions.ts`
so #2493 can consume the same parser.

Decisions opt out of both gates via `### Claude's Discretion` heading
or `[informational]` / `[folded]` / `[deferred]` tags.

Both gates skip silently when `workflow.context_coverage_gate=false`
(default `true`).

Closes #2492

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

* fix(#2492): make plan-phase decision gate actually block (review F1, F8, F9, F10, F15)

- F1: replace `${context_path}` with `${CONTEXT_PATH}` in the plan-phase
  gate snippet so the BLOCKING gate receives a non-empty path. The
  variable was defined in Step 4 (`CONTEXT_PATH=$(_gsd_field "$INIT" ...)`)
  and the gate snippet referenced the lowercase form, leaving the gate to
  run with an empty path argument and silently skip.
- F15: wrap the SDK call with `jq -e '.data.passed == true' || exit 1` so
  failure halts the workflow instead of being printed and ignored. The
  verify-phase counterpart deliberately keeps no exit-1 (non-blocking by
  design) and now carries an inline note documenting the asymmetry.
- F10: tag the JSON example fence as `json` and the options-list fence as
  `text` (MD040).
- F8/F9: anchor the heading-presence test regexes to `^## 13[a-z]?\\.` so
  prose substrings like "Requirements Coverage Gate" mentioned in body
  text cannot satisfy the assertion. Added two new regression tests
  (variable-name match, exit-1 guard) so a future revert is caught.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

* fix(#2492): tighten decision-coverage gates against false positives and config drift (review F3,F4,F5,F6,F7,F16,F18,F19)

- F3: forward `workstream` arg through both gate handlers so workstream-scoped
  `workflow.context_coverage_gate=false` actually skips. Added negative test
  that creates a workstream config disabling the gate while the root config
  has it enabled and asserts the workstream call is skipped.
- F4: restrict the plan-phase haystack to designated sections — front-matter
  `must_haves` / `truths` / `objective` plus body sections under headings
  matching `must_haves|truths|tasks|objective`. HTML comments and fenced
  code blocks are stripped before extraction so a commented-out citation or
  a literal example never counts as coverage. Verify-phase keeps the broader
  artifact-wide haystack by design (non-blocking).
- F5: reject decisions with fewer than 6 normalized words from soft-matching
  (previously only rejected when the resulting phrase was under 12 chars
  AFTER slicing — too lenient). Short decisions now require an explicit
  `D-NN` citation, with regression tests for the boundary.
- F6: walk every `*-SUMMARY.md` independently and use `matchAll` with the
  `/g` flag so multiple `files_modified:` blocks across multiple summaries
  are all aggregated. Previously only the first block in the concatenated
  string was parsed, silently dropping later plans' files.
- F7: validate every `files_modified` path stays inside `projectDir` after
  resolution (rejects absolute paths, `../` traversal). Cap each file read
  at 256 KB. Skipped paths emit a stderr warning naming the entry.
- F16: validate `workflow.context_coverage_gate` is boolean in
  `loadGateConfig`; warn loudly on numeric or other-shaped values and
  default to ON. Mirrors the schema-vs-loadConfig validation gap from
  #2609.
- F18: bump verify-phase `git log -n` cap from 50 to 200 so longer-running
  phases are not undercounted. Documented as a precision-vs-recall tradeoff
  appropriate for a non-blocking gate.
- F19: tighten `QueryResult` / `QueryHandler` to be parameterized
  (`<T = unknown>`). Drops the `as unknown as Record<string, unknown>`
  casts in the gate handlers and surfaces shape mismatches at compile time
  for callers that pass a typed `data` value.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

* fix(#2492): harden decisions parser and verify-phase glob (review F11,F12,F13,F14,F17,F20)

- F11: strip fenced code blocks from CONTEXT.md before searching for
  `<decisions>` so an example block inside ``` ``` is not mis-parsed.
- F12: accept tab-indented continuation lines (previously required a leading
  space) so decisions split with `\t` continue cleanly.
- F13: parse EVERY `<decisions>` block in the file via `matchAll`, not just
  the first. CONTEXT.md may legitimately carry more than one block.
- F14: `decisions.parse` handler now resolves a relative path against
  `projectDir` — symmetric with the gate handlers — and still accepts
  absolute paths.
- F17: replace `ls "${PHASE_DIR}"/*-CONTEXT.md | head -1` in verify-phase.md
  with a glob loop (ShellCheck SC2012 fix). Also avoids spawning an extra
  subprocess and survives filenames with whitespace.
- F20: extend the unicode quote-stripping in the discretion-heading match
  to cover U+2018/2019/201A/201B and the U+201C-F double-quote variants
  plus backtick, so any rendering of "Claude's Discretion" collapses to
  the same key.

Each fix has a regression test in `decisions.test.ts`.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This commit is contained in:
Tom Boucher
2026-04-23 00:26:53 -04:00
committed by GitHub
parent 1a3d953767
commit f30da8326a
16 changed files with 1900 additions and 4 deletions

View File

@@ -1310,6 +1310,72 @@ Options:
If `TEXT_MODE` is true, present as a plain-text numbered list (options already shown in the block above). Otherwise use AskUserQuestion to present the options.
## 13a. Decision Coverage Gate
After the requirements coverage gate passes, verify that every trackable
decision captured by discuss-phase in CONTEXT.md `<decisions>` is referenced
by at least one plan. This is the **translation gate** from issue #2492 —
its job is to refuse to mark a phase planned when a discuss-phase decision
silently dropped on the way into the plans.
**Skip if** `workflow.context_coverage_gate` is explicitly set to `false`
(absent key = enabled). Also skip if no CONTEXT.md exists for this phase
(nothing to translate) or if its `<decisions>` block is empty.
```bash
GATE_CFG=$(gsd-sdk query config-get workflow.context_coverage_gate 2>/dev/null || echo "true")
if [ "$GATE_CFG" != "false" ]; then
GATE_RESULT=$(gsd-sdk query check.decision-coverage-plan "${PHASE_DIR}" "${CONTEXT_PATH}")
# BLOCKING: refuse to mark phase planned when a trackable decision is uncovered.
# `passed: true` covers both real-pass and skipped cases (gate disabled / no CONTEXT.md /
# no trackable decisions). Verify-phase counterpart deliberately omits this exit-1 — that
# gate is non-blocking by design (review finding F15).
echo "$GATE_RESULT" | jq -e '.data.passed == true' >/dev/null || {
echo "$GATE_RESULT" | jq -r '.data.message'
exit 1
}
fi
```
The handler returns JSON:
```json
{
"passed": true,
"skipped": false,
"total": 2,
"covered": 2,
"uncovered": [ { "id": "D-01", "text": "...", "category": "..." } ],
"message": "..."
}
```
**If `passed` is true (or `skipped` is true):** Display
`✓ Decision coverage: {M}/{N} CONTEXT.md decisions covered by plans` (or
`(skipped — gate disabled)` / `(skipped — no decisions)`) and proceed to
step 13b.
**If `passed` is false:** Display the handler's `message` block. It already
names each uncovered decision (`D-NN | category | text`) and tells the user
what to do — cite the id in a relevant plan's `must_haves` / `truths`, or
move the decision under `### Claude's Discretion` / tag it `[informational]`
if it should not be tracked. Then offer:
```text
Options:
1. Re-plan to cover missing decisions (recommended)
2. Edit CONTEXT.md to mark dropped decisions as [informational] / Discretion
3. Proceed anyway — accept the coverage gap
```
If `TEXT_MODE` is true, present as a plain-text numbered list. Otherwise use
AskUserQuestion. Selecting "Proceed anyway" continues to step 13b but
records the override in STATE.md so verify-phase can re-surface it.
**Why this gate blocks:** failing here is cheap. The plans are the contract
between discuss-phase and execute-phase; if a decision isn't visible in any
plan, no executor will implement it. Catching that now beats discovering it
after thousands of dollars of execution.
## 13b. Record Planning Completion in STATE.md
After plans pass all gates, record that planning is complete so STATE.md reflects the new phase status:

View File

@@ -183,6 +183,57 @@ grep -E "Phase ${PHASE_NUM}" .planning/REQUIREMENTS.md 2>/dev/null || true
For each requirement: parse description → identify supporting truths/artifacts → status: ✓ SATISFIED / ✗ BLOCKED / ? NEEDS HUMAN.
</step>
<step name="verify_decisions">
**Decision coverage validation gate (issue #2492).**
After requirements coverage, also check that each trackable CONTEXT.md
`<decisions>` entry shows up somewhere in the shipped artifacts (plans,
SUMMARY.md, files modified by the phase, or recent commit subjects on the
phase branch).
This gate is **non-blocking / warning only** by deliberate asymmetry with
the plan-phase translation gate. The plan-phase gate already blocked at
translation time, so by the time verification runs every decision has
either been translated or explicitly deferred. This gate's job is to
surface decisions that *were* translated but vanished during execution —
that's a soft signal because "honors a decision" is a fuzzy substring
heuristic, and we don't want a paraphrase miss to fail an otherwise good
phase.
**Skip if** `workflow.context_coverage_gate` is explicitly set to `false`
(absent key = enabled). Also skip cleanly when CONTEXT.md is missing or has
no `<decisions>` block.
```bash
GATE_CFG=$(gsd-sdk query config-get workflow.context_coverage_gate 2>/dev/null || echo "true")
if [ "$GATE_CFG" != "false" ]; then
# Discover the phase CONTEXT.md via glob expansion rather than `ls | head`
# (review F17 / ShellCheck SC2012). Globs preserve filenames containing
# spaces and avoid an extra subprocess.
CONTEXT_PATH=""
for f in "${PHASE_DIR}"/*-CONTEXT.md; do
[ -e "$f" ] && CONTEXT_PATH="$f" && break
done
DECISION_RESULT=$(gsd-sdk query check.decision-coverage-verify "${PHASE_DIR}" "${CONTEXT_PATH}")
fi
```
The handler returns JSON `{ skipped, blocking: false, total, honored,
not_honored: [...], message }`.
**Reporting:** Append the handler's `message` (a `### Decision Coverage`
section) to VERIFICATION.md regardless of outcome — even when all
decisions are honored, recording the count helps reviewers spot drift over
time. Set `decision_coverage` in the verification result to
`{honored, total, not_honored: [...]}` so downstream tooling can read it.
**Status impact:** none. The decision gate does NOT influence the
`gaps_found` / `human_needed` / `passed` decision tree in
`determine_status`. Its findings are warnings the user reviews and may act
on by re-opening the phase or by acknowledging the decision was abandoned
intentionally.
</step>
<step name="behavioral_verification">
**Run the project's test suite and CLI commands to verify behavior, not just structure.**
@@ -479,6 +530,7 @@ Orchestrator routes: `passed` → update_roadmap | `gaps_found` → create/execu
- [ ] All artifacts checked at all three levels
- [ ] All key links verified
- [ ] Requirements coverage assessed (if applicable)
- [ ] CONTEXT.md decisions checked against shipped artifacts (#2492 — non-blocking)
- [ ] Anti-patterns scanned and categorized
- [ ] Test quality audited (disabled tests, circular patterns, assertion strength, provenance)
- [ ] Human verification items identified