abilityai/quick-search
.claude/skills/quick-search/SKILL.md
Fast knowledge base retrieval - single FAISS search + graph connections, no subagents or LLM orchestration
$ install --global
skillsauthnpx skillsauth add abilityai/cornelius quick-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: Mar 28, 2026, 2:18 AM36.8s1 file scanned
SKILL.md
- name:
- quick-search
- description:
- Fast knowledge base retrieval - single FAISS search + graph connections, no subagents or LLM orchestration
- argument-hint:
- <search query>
- allowed-tools:
- [Bash, Read]
- user-invocable:
- true
Quick Search
Fastest possible knowledge base retrieval. No subagents, no multi-layer orchestration, no changelogs.
Purpose
Return relevant notes and their graph neighborhood in minimal tool calls. Designed for speed over thoroughness.
Query
$ARGUMENTS
Process
Execute these two commands in parallel (single message, two Bash calls):
# 1. Semantic search (6-14s - the unavoidable cost)
/Users/eugene/Dropbox/Agents/Cornelius/resources/local-brain-search/run_search.sh "$ARGUMENTS" --limit 5 --json
# 2. Graph connections for likely top hit (0.3s - nearly free)
/Users/eugene/Dropbox/Agents/Cornelius/resources/local-brain-search/run_connections.sh "$ARGUMENTS" --json
Then read the top result file using Read tool.
That's it. Three tool calls. Present results and stop.
Output Format
Keep it brief:
## [Query]
**Top matches:**
1. [[Note Title]] (0.XX) - [one-line summary from content]
2. [[Note Title]] (0.XX) - [one-line summary]
3. [[Note Title]] (0.XX) - [one-line summary]
**Graph neighborhood** (for top hit):
- Outgoing: [[Note]], [[Note]], ...
- Incoming: [[Note]], [[Note]], ...
**Top result content:**
[First ~30 lines of the highest-scoring note]
Rules
- NO subagent spawning - do everything inline
- NO changelog creation - this is a read-only lookup
- NO multi-layer expansion - one search, one connection call, done
- NO spreading activation - use static mode (faster for simple lookups)
- Parallel execution - run search and connections in the same message
- Maximum 3 tool calls - search + connections + read top file
- If the user wants deeper analysis, tell them to use
/recallor/find-connections
Related Skills
abilityai/update-voice-prompt
development
VerifiedTrustedCommunity
Rebuild the Cornelius voice agent system prompt from knowledge base sources
78SKILL.mdUpdated Mar 27, 2026
abilityai/update-dashboard
data-ai
VerifiedTrustedCommunity
Update dashboard.yaml with current knowledge base metrics from analysis report
78SKILL.mdUpdated Mar 27, 2026
abilityai/update-changelog
documentation
VerifiedTrustedCommunity
Update Knowledge Graph Changelog
78SKILL.mdUpdated Mar 27, 2026
abilityai/test-memory-system
testing
VerifiedTrustedCommunity
Comprehensive testing playbook for Local Brain Search memory improvements (Phases 1, 3, 4)
78SKILL.mdUpdated Mar 27, 2026