phase-10i-transcript-recovery-final-re-review — Claude Code prompt
phase-10i-transcript-recovery-final-re-review is a Claude Code prompt for re-reviewing transcript recovery and verifying attachment metadata redaction fixes.
Apache-2.0 ★ 1,497 updated 2026-06-16
.claude/commands/phase-10i-transcript-recovery-final-re-review.mdWhat phase-10i-transcript-recovery-final-re-review does
This prompt instructs the model to act as a principal engineer and architect conducting a narrow re-review of transcript recovery logic. Its main target is checking whether an attachment metadata redaction bypass issue was fixed without introducing fresh P0, P1, or P2 regressions.
Execution relies on reading local project files, inspecting recent diffs, and running verification steps. Specifically, it executes bun test across six test suites alongside bun run lint and bun run typecheck. The output goes directly into a dedicated Markdown findings report, while modifying source files or committing changes is explicitly prohibited.
Because file paths in the prompt reference hardcoded local directories under /Users/truffle/work/..., using this workflow outside its target project requires updating the paths and test file lists.
What it helps with
- verify that attachment metadata redaction issues are resolved
- check targeted code changes for new regressions across critical levels
- run targeted Bun test suites, linter, and type checks
- produce an inspection report without modifying production code
Example prompts for the agent
- Run the phase 10I final re-review command from .claude/commands/phase-10i-transcript-recovery-final-re-review.md
- Verify the redaction fix against the test suite and save the audit report
How to install phase-10i-transcript-recovery-final-re-review
Manually: put the files in place
.claude/commands/phase-10i-transcript-recovery-final-re-review.mdHow to use
Paste the text at the start of a chat or into project instructions
Questions about phase-10i-transcript-recovery-final-re-review
How to install phase-10i-transcript-recovery-final-re-review in Claude Code?
Place the file at .claude/commands/phase-10i-transcript-recovery-final-re-review.md inside your repository.
What does phase-10i-transcript-recovery-final-re-review do?
It guides Claude Code through reviewing redaction and transcript search files, running Bun verification suites, and generating an audit report.
Can the assistant edit source code when using this prompt?
No, the prompt explicitly instructs the assistant not to edit production files and not to make git commits.
Source text
phase-10i-transcript-recovery-final-re-review.md · Apache-2.0
ultrathink. ultrathink. ultrathink. You are a principal engineer, principal architect, and principal product manager at Anthropic, all three at once. You are doing a final narrow re-review of Phase 10I transcript recovery after the previous re-review found one remaining P2: attachment metadata bypassed redaction. No context anxiety. No token limits. No time pressure. No cost anxiety. No "v2" thinking. There is no v2. Build it right. Take as long as you need. ## Mission Verify whether the attachment metadata redaction P2 is now resolved and whether the narrow fix introduced any new P0, P1, or P2 issue. ## Required Reading Read from disk: 1. `/Users/truffle/work/phantom-murph-hardening/research/chat-experience/phase-10i-transcript-recovery-re-review.md` 2. `src/chat/transcript-search.ts` 3. `src/chat/redaction.ts` 4. `src/chat/__tests__/transcript-search.test.ts` 5. `src/agent/__tests__/reflective-transcript-tools.test.ts` Also inspect the current diff for related files. ## Required Verification Run: ```bash bun test src/chat/__tests__/transcript-search.test.ts src/chat/__tests__/continuity-context.test.ts src/agent/__tests__/reflective-transcript-tools.test.ts src/chat/__tests__/run-timeline.test.ts src/agent/__tests__/agent-sdk-boundary-callers.test.ts src/agent/__tests__/murph-context.test.ts bun run lint bun run typecheck ``` ## Output Contract Write your final report to: `/Users/truffle/work/phantom-murph-hardening/research/chat-experience/phase-10i-transcript-recovery-final-re-review.md` Lead with findings. If there are no P0, P1, or P2 findings, say so explicitly. Do not edit production code. Do not commit. ultrathink before finalizing. Actual files and command output are evidence.
© ghostwright · Apache-2.0 · ghostwright/phantom