jimweller/transcript-search
dotfiles/claude-code/skills/transcript-search/SKILL.md
Search saved session transcripts for past decisions, actions, errors, and context that has left the current conversation window.
testing
Updated Apr 19, 2026
$ install --global
skillsauthnpx skillsauth add jimweller/dotfiles transcript-searchInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: Apr 19, 2026, 5:01 AM4.5s1 file scanned
SKILL.md
- name:
- transcript-search
- description:
- Search saved session transcripts for past decisions, actions, errors, and context that has left the current conversation window.
<!-- markdownlint-disable-file MD041 -->
STARTER_CHARACTER = 🗃️
Transcript Search
Search the global session memory database for past conversations using hybrid FTS5 + vector retrieval.
When to use
- User asks about something from a previous session or earlier in a long session
- User references a past decision, error, action, or discussion no longer in context
- User asks "what did we do about X", "when did we fix Y", "what error did we get"
Database location
~/.claude/session_memory.db (global, all projects)
Retrieval tool
TR=~/.config/dotfiles/dotfiles/claude-code/tools/total-recall
$TR/.venv/bin/python $TR/retrieve.py "QUERY" --db ~/.claude/session_memory.db
Default output is --format context which produces XML-tagged conversation
excerpts ready to read directly.
Common flags
| Flag | Purpose |
| ------------------------ | ---------------------------------------------------- |
| --project NAME | Restrict to a specific project |
| --top-k N | Number of seed matches before expansion (default: 5) |
| --window N | Context window +/- N around each match (default: 2) |
| --budget N | Max chunks in final output (default: 20) |
| --tokens N | Max tokens in final output (default: 6000) |
| --session ID | Restrict to a specific session_id |
| --no-tools | Exclude tool_use/tool_result chunks |
| --no-vectors | FTS5 keyword search only (skip embeddings) |
| --format json | JSON array output for programmatic use |
| --roles user,assistant | Filter by role |
Search procedure
- Run retrieve.py with the user's query
- Read the returned context block
- If results are too broad, narrow with
--projector--session - If results are too sparse, increase
--top-kor--budget - For keyword-exact searches, use
--no-vectorsto restrict to FTS5
Tips
- The retrieval pipeline uses RRF fusion of keyword + vector search, window expansion, ancestor backtracking, and cross-session semantic links
- Use
--no-toolswhen searching for decisions or discussion (skips noisy tool output) - Use
--format jsonwhen you need to process results programmatically
Related Skills
jimweller/transcript-search
testing
VerifiedTrustedCommunity
Search saved session transcripts for past decisions, actions, errors, and context that has left the current conversation window.
1SKILL.mdUpdated Apr 22, 2026
jimweller/prd-review
data-ai
VerifiedTrustedCommunity
Review a PRD for defects via Claude opus subagent.
SKILL.mdUpdated Apr 29, 2026
jimweller/md-syntax
development
VerifiedTrustedCommunity
Markdown authoring guidelines for formatting, code blocks, and structure. Use when writing or editing markdown files.
SKILL.mdUpdated Apr 29, 2026
jimweller/configs/claude-code/skills/md-style
development
VerifiedTrustedCommunity
--- name: md-style description: README style guide for concise, direct documentation. Use when writing or editing README files. ß--- <!-- markdownlint-disable-file MD041 --> # README Style Guide Write concise, direct README files for experienced engineers. ## Principles - **No fluff** - Skip tables of contents, verbose explanations, development history - **No roadmaps** - Document current state only, not plans or decisions. Readme is an engineering specification. Not a project plan or chan
SKILL.mdUpdated Apr 29, 2026