docs(openspec): archive feedback-channels

[theCodeDrift]

Stack (root → tip):

• docs(openspec): propose feedback-channels #477
  • feat(feedback): key surveys by kind and discriminate the payload on it #478
    • feat(feedback): send each kind to its survey, fill a bug's version info, and point to issues when telemetry is off #479
      • feat(agent): add the feedback and bug-report recipes, and rename the invited survey's recipe to rule-feedback #480
        • feat(skill): route feedback and bug reports to their recipes #481
          • ➡️ docs(openspec): archive feedback-channels #482 (you are here)

Archives the change into openspec/specs/cli-feedback-survey/spec.md.

Scenario check, comparing the spec before and after the archive:

• The only scenario removed is "The feedback recipe embeds the schema", which returns renamed as "The rule-feedback recipe embeds the schema".
• Every scenario in the delta is present afterwards.

Task 5.3, a live send to PostHog, was closed without sending. The survey and question ids come from the live survey definitions, the rule survey's content is reused unchanged, and the tests pin each kind's keys.

This stack lands with gh stack merge, which merges every PR to main in one all-or-nothing operation, so no layer reaches main on its own. That matters because the layers only work together. For example, #478 requires kind on the payload while the recipe the invite names still describes a single survey, and #480 lists bug-report in the index before the skill routes to it.

[theCodeDrift]

@claude /review

[github-actions]

Claude finished @theCodeDrift's task in 1m 15s —— View job

---

Review mode: incremental — read 0 prior review thread(s) before reviewing.

.prior-review.json contained only the @claude /review trigger comment itself — no reviews and no review threads. This is effectively the first review of this PR; nothing to carry forward as "already addressed."

• Read .prior-review.json (prior review data)
• Fetch PR diff and metadata (gh pr diff, gh pr view --json commits,headRefOid,files)
• Review changed files for correctness/clarity
• Post findings

What this PR does

Archives the feedback-channels OpenSpec change: renames openspec/changes/feedback-channels/ → openspec/changes/archive/2026-10-08-feedback-channels/, marks tasks 5.1/5.2 done and documents 5.3 as intentionally skipped (sends real payloads into production PostHog surveys, held for explicit go-ahead), and rewrites openspec/specs/cli-feedback-survey/spec.md to the post-archive result.

Verification performed

Given this repo's CLAUDE.md explicitly warns that openspec archive silently replaces a requirement with whatever a delta's MODIFIED/ADDED/REMOVED blocks contain — so a dropped scenario or a renamed requirement treated as MODIFIED instead of REMOVED+ADDED loses content with no error — I diffed the archived change's delta spec (openspec/changes/archive/2026-10-08-feedback-channels/specs/cli-feedback-survey/spec.md) against the resulting standing spec (openspec/specs/cli-feedback-survey/spec.md) line by line rather than trusting the PR description's self-check:

• The untouched requirement ("A survey invite is appended to surveyed recipes…") — not mentioned in the delta — is carried over verbatim. Correct.
• MODIFIED blocks for "The cadence store…", "The feedback subcommand…", and "The feedback payload uses human keys…" each match by requirement title and land with every prior scenario preserved plus the new ones. Correct.
• The rename case is handled properly as REMOVED ("The feedback and feedback-invite recipes") + ADDED ("The rule-feedback and feedback-invite recipes") rather than a MODIFIED — which is exactly the trap CLAUDE.md calls out, since a same-named MODIFIED would have silently matched nothing. The removed requirement's content survives intact under the new title/name.
• The three newly ADDED requirements ("The general feedback recipe", "The bug-report recipe", "The CLI supplies a bug report's version information") land exactly as authored.
• No scenario present before the archive is missing after it, other than the one the PR description names as renamed ("The feedback recipe embeds the schema" → "The rule-feedback recipe embeds the schema"), which I confirmed independently.

The tasks.md rename/edit and the four pure renames (.openspec.yaml, design.md, proposal.md, delta spec.md) carry 0 content diff, as expected for an archive move.

Findings

Found nothing new. This is a mechanical, well-verified archive — the result matches what the delta specifies, the one requirement rename is handled with the correct REMOVED+ADDED pair rather than a silently-matched MODIFIED, and the deferred task (5.3, sending real PostHog payloads) is appropriately left unchecked with its reason documented.