pascalandy/delegate-to-sub
dot_config/ai_templates/skills/devtools/delegate-to-sub/SKILL.md
Explicitly triggered when the user mentions `delegate-to-sub`. It routes work to sub-agents.
testing
Updated Apr 24, 2026
$ install --global
skillsauthnpx skillsauth add pascalandy/dotfiles delegate-to-subInstall 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 24, 2026, 8:34 PM0.4s1 file scanned
SKILL.md
- name:
- delegate-to-sub
- description:
- >
Using Sub-Agents
When the main agent (1-kimi) needs to delegate work, use this skill to determine which sub-agent to call.
Matrix
-
1-kimi(Kimi 2p6 via opencode-go)- Speed: Fastest | Cost: Lowest | Intelligence: High
- Use for: Default for everything. First-pass reviews, commits, quick edits, general tasks, exploration, research.
-
2-gpt(GPT-5.4 High Reasoning via OpenAI)- Speed: Medium | Cost: Medium-high | Intelligence: Very high
- Use for: Second-pass reviews. Complex reasoning, different perspective from Kimi.
-
3-minimax(MiniMax 2.7 via opencode-go)- Speed: Medium-fast | Cost: Low | Intelligence: High
- Use for: Third-pass reviews. Alternative perspective, batch tasks, bulk operations.
-
gpthigh(GPT-5.4 High Reasoning via OpenAI)- Speed: Medium | Cost: Medium-high | Intelligence: Very high
- Use for: Alternative deep reasoning when GPT-5.4 is already in use.
-
gemini(Gemini 3.1 Pro via OpenRouter)- Speed: Medium | Cost: Medium | Intelligence: High
- Use for: Fourth-pass reviews only when user explicitly requests 4 rounds.
-
glm(GLM 5.1 via zai-coding-plan)- Speed: Medium-fast | Cost: Low | Intelligence: High (coding-optimized)
- Use for: Batch tasks, bulk operations, large-scale refactoring, heavy tasks where cost matters.
Review Workflow
When the user requests a review, they explicitly say "do a two pass review", "do a three pass review", or "do a four pass review". The main agent (1-kimi) spawns sub-agents accordingly:
2-Pass Review:
- Main agent calls subagent
1-kimi(fresh instance) for initial review - Main agent calls subagent
2-gptfor thorough, deep analysis
3-Pass Review (most common):
- Main agent calls subagent
1-kimi(fresh instance) for initial review - Main agent calls subagent
2-gptfor thorough, deep analysis - Main agent calls subagent
3-minimaxfor alternative perspective
4-Pass Review (when user explicitly requests):
- Main agent calls subagent
1-kimi(fresh instance) for initial review - Main agent calls subagent
2-gptfor thorough, deep analysis - Main agent calls subagent
3-minimaxfor alternative perspective - Main agent calls subagent
geminifor additional perspective
Related Skills
pascalandy/wiki-map
development
VerifiedTrustedCommunity
Explicitly triggered when the user mentions `wiki-map`. It ingests, queries, lints, or compiles a markdown wiki.
SKILL.mdUpdated Apr 26, 2026
pascalandy/single-skill-creator
testing
VerifiedTrustedCommunity
Explicitly triggered when the user mentions `single-skill-creator`. It scaffolds a new `SKILL.md`.
SKILL.mdUpdated Apr 26, 2026
pascalandy/qmd
testing
VerifiedTrustedCommunity
Explicitly triggered when the user mentions `qmd`. It searches local markdown or QMD collections.
SKILL.mdUpdated Apr 26, 2026
pascalandy/ontology-map
development
VerifiedTrustedCommunity
Explicitly triggered when the user mentions `ontology-map`. It builds, refreshes, or checks an ontology.
SKILL.mdUpdated Apr 26, 2026