This release includes 1 breaking change for platform teams planning a safe upgrade.
✓ No known CVEs patched in this version
Summary
AI summaryRemoved placeholder telemetry line from .env.example.
Full changelog
[1.3.2] — 2026-04-24
Closes the audit loop opened by ship-check during the 1.3.1 retrospective. Pure docs + metadata + skill-library hygiene; no engine code changes.
Added
server.jsonenv-var declarations for the search-enrichment feature —SEARCH_PROVIDER,SEARCH_API_KEY(secret),SEARCH_API_URL. These were documented in.env.examplesince the optional context-enrichment feature shipped in 1.1.x but never made it into the MCP Registry manifest. Users installing via the registry UI will now be prompted for them.- README env-var reference table rows for
SEARCH_*— same gap on the docs side, also closed. - Synced
server.jsondescription to match the 1.3.0+ pitch ("persistent memory, knowledge packs, explicit token-budget curation…"). Was still on 1.2.0-era language.
Changed
- Removed the
CLARIFYPROMPT_TELEMETRYplaceholder line from.env.example. It was a forward-looking commented-out example for a feature that doesn't exist yet. Documenting features that don't exist makes the docs slightly lie. Will reappear the moment any telemetry code lands.
Skill library — consolidation pass
The ship-check skill caught its own first audit findings during the 1.3.1 → 1.3.2 cycle. Used the moment to run the first promotion pass under the cascade-learning model we established:
- Promoted CP-7 → general check #9 (License consistency + commitment-drift heuristic). Now applies to any project regardless of language.
- Promoted CP-9 → general check #10 (Build-artifact directory hygiene across Node / Python / Rust / Java / Go).
- Promoted CP-2 + CP-3 → general check #11 (MCP server-manifest env-var cross-check, conditional on
server.jsonexistence — now also reconciles against the README env-var table). - Audit-logic fixes in the user-scoped skill: secrets sweep now excludes
.claude/skills/**(skill files commonly enumerate token patterns; matching them is a self-reference false positive);.env.exampleparser now includes commented# VAR=lines.
The project-scoped skill at .claude/skills/ship-check/SKILL.md now carries a promotion log — a dated audit trail of which checks moved from project-scoped to user-scoped, with one-line rationales each.
Notes for integrators
- npm tarball still ships
dist/+packs/+README.md+LICENSE+CHANGELOG.md+.env.example. No file-list changes. - No new MCP tools (still 16 from 1.3.0).
- No env vars removed (only the
CLARIFYPROMPT_TELEMETRYexample comment was removed, and that was never a live env var).
Breaking Changes
- Removed the `CLARIFYPROMPT_TELEMETRY` placeholder line from `.env.example`.
Weekly OSS security release digest.
The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.
No spam, unsubscribe anytime.
Share this release
About LumabyteCo/clarifyprompt-mcp
MCP server for AI prompt optimization — transforms vague prompts into platform-optimized prompts for 58+ AI platforms across 7 categories (image, video, voice, music, code, chat, document).
Related context
Beta — feedback welcome: [email protected]