shynlee04/hivemind-plugin/hm-l2-lineage-router
Claude Code and agent skill by shynlee04/hivemind-plugin
- community
Checking that the SKILL.md file is still in the repository.
Route tasks to correct hm-* skills based on intent classification. Maps task categories to skill loading bundles. Use when an agent needs to determine which hm-* skills to load for a given task. Triggers: "route task", "classify intent", "which skill to load", "task routing", "lineage routing", "skill bundle for task", "load skills for", "what skills do I need", "skill loading order", "determine skill lineage", "map task to skills", "skill selection for workflow". NOT for executing skills — only for determining which skills to load. Max 5 skills per bundle.
Install
npx vskill@latest install shynlee04/hivemind-plugin/hm-l2-lineage-routerSKILL.md
The Iron Law
Lineage Router
Overview
Given a task intent, determine which hm-* skills should be loaded. This skill maps task categories to skill loading bundles. It does NOT execute skills — it only determines which skills to load and in what order.
Six task categories, six skill bundles:
| Category | Skills | Max | |
|---|---|---|---|
| Research | hm-detective + hm-deep-research + hm-tech-stack-ingest | 3 | |
| Planning | hm-spec-driven-authoring + hm-planning-persistence | 2 | |
| Execution | hm-phase-execution + hm-cross-cutting-change | 2 | |
| Quality | hm-test-driven-execution + hm-gate-orchestrator | 2 | |
| Debug | hm-debug + hm-completion-looping | 2 | |
| Review | hm-production-readiness + hm-gate-orchestrator | 2 |
On Load
- Read
references/routing-map.md— the complete routing map with priority ordering and dependency chains - Identify the task intent from the current context
- Classify the task into one of the six categories
Trigger Phrases
- "route task" / "route this task"
- "classify intent" / "classify task intent"
- "which skill to load" / "what skills do I need"
- "task routing" / "lineage routing"
- "skill bundle for task" / "load skills for"
- "skill loading order" / "determine skill lineage"
- "map task to skills" / "skill selection for workflow"
Routing Map
Category 1: Research Tasks
Intent signals: "investigate", "research", "find out", "analyze", "look into", "explore"
| Priority | Skill | Role | |
|---|---|---|---|
| 1 | hm-detective | Codebase investigation (SCAN/READ/DEEP modes) | |
| 2 | hm-deep-research | Version-matched deep research with citations | |
| 3 | hm-tech-stack-ingest | Cache third-party docs for offline reference |
Loading order: hm-tech-stack-ingest (if stack cache needed) → hm-detective → hm-deep-research
Category 2: Planning Tasks
Intent signals: "plan", "spec", "requirements", "design", "architect"
| Priority | Skill | Role | |
|---|---|---|---|
| 1 | hm-spec-driven-authoring | Falsifiable requirements + acceptance criteria | |
| 2 | hm-planning-persistence | 3-file external memory for multi-session planning |
Loading order: hm-planning-persistence (setup) → hm-spec-driven-authoring (content)
Category 3: Execution Tasks
Intent signals: "implement", "build", "execute", "run the phase", "code"
| Priority | Skill | Role | |
|---|---|---|---|
| 1 | hm-phase-execution | Wave-based parallelization + checkpoint recovery | |
| 2 | hm-cross-cutting-change | Cross-pane modification safety for breaking changes |
Loading order: hm-phase-execution (primary) → hm-cross-cutting-change (when touching multiple layers)
Category 4: Quality Tasks
Intent signals: "test", "verify", "quality", "validate", "TDD", "red-green-refactor"
| Priority | Skill | Role | |
|---|---|---|---|
| 1 | hm-test-driven-execution | RED/GREEN/REFACTOR cycles with runtime truth | |
| 2 | hm-gate-orchestrator | Quality gate triad pipeline (lifecycle → spec → evidence) |
Loading order: hm-test-driven-execution (implementation) → hm-gate-orchestrator (validation)
Category 5: Debug Tasks
Intent signals: "debug", "fix", "broken", "failing", "error", "investigate issue"
| Priority | Skill | Role | |
|---|---|---|---|
| 1 | hm-debug | Systematic debugging with persistent state | |
| 2 | hm-completion-looping | Guardrail against premature completion claims |
Loading order: hm-debug (investigation) → hm-completion-looping (verification)
Category 6: Review Tasks
Intent signals: "review", "audit", "readiness", "deploy check", "ship ready"
| Priority | Skill | Role | |
|---|---|---|---|
| 1 | hm-production-readiness | Deployment safety verification (8 dimensions) | |
| 2 | hm-gate-orchestrator | Quality gate triad for final approval |
Loading order: hm-production-readiness (evidence collection) → hm-gate-orchestrator (gate verification)
Loading Rules
- Max 5 skills per bundle. If a task needs more than 5, split the task.
- Load in priority order. Lower-numbered skills load first.
- Respect dependencies. If Skill B depends on Skill A's output, load A first.
- Classify before loading. Determine the category before loading any skills.
- Multi-category tasks. If a task spans 2 categories, load the primary category's bundle first, then add 1-2 skills from the secondary category (staying within the 5-skill limit).
Multi-Category Decision Tree
Self-Correction
Anti-Pattern 1: Overloading
Detection: More than 5 skills being loaded for a single task. Correction: Split the task into subtasks. Each subtask gets its own bundle. Signal: "Task too broad — split into [A] and [B], then route each independently."
Anti-Pattern 2: Wrong Lineage
Detection: Loading hf- skills for an hm- task, or vice versa. Correction: This router handles hm-* lineage only. If the task is a meta-builder task (creating skills, agents, commands), route to hf-meta-builder instead. If the task is a product-dev task, use this router.
Anti-Pattern 3: Missing Input
Detection: Task intent cannot be classified into any of the 6 categories. Correction: If the task is genuinely new (e.g., "monitor production"), route to the closest category and flag: "No exact match. Routed to [closest category]. Verify the bundle covers the task's needs."
Anti-Pattern 4: Stale Bundles
Detection: A loaded skill references another skill not in the current bundle, and the referenced skill is critical for the task. Correction: Add the missing skill to the bundle if the total stays ≤ 5. If adding would exceed 5, remove the lowest-priority skill from the current bundle and document the trade-off.
Quality Contract (HMQUAL Compliance)
| HMQUAL | Compliance | Evidence | |
|---|---|---|---|
| HMQUAL-01 | Trigger phrases ≥7 in description | 12 trigger phrases in description | |
| HMQUAL-02 | Self-correction with 4 anti-patterns | Overloading, Wrong Lineage, Missing Input, Stale Bundles | |
| HMQUAL-03 | Cross-references to sibling skills | 13 skill references across 6 bundles + 4 consumer skills | |
| HMQUAL-04 | Progressive disclosure | SKILL.md (routing map) + references/routing-map.md (detailed routing) | |
| HMQUAL-05 | Evals with 3 scenarios | evals/evals.json — 3 trigger scenarios | |
| HMQUAL-06 | Metrics scorecard | metrics/rich-gate-scorecard.md | |
| HMQUAL-07 | Iron law enforcement | "Max 5 skills per bundle." | |
| HMQUAL-08 | Honest RICH scoring | Scored in metrics/rich-gate-scorecard.md |
Cross-References
| Related Skill | Boundary | |
|---|---|---|
hm-coordinating-loop | Consumer — uses this router to determine skill bundles for delegated tasks | |
hm-phase-execution | Consumer — uses this router to determine execution-time skill loading | |
hm-phase-loop | Consumer — uses this router for iterative phase skill selection | |
hm-subagent-delegation-patterns | Consumer — uses this router to attach skill bundles to delegation packets | |
hm-gate-orchestrator | Referenced in Quality and Review bundles — the gate triad orchestrator |
Security
- Security scanPassed, 91 of 100Pattern checks on the skill's files.
- Model reviewNot runA model review is scheduled for the next rescan.
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingassets/.hivemind/skills/hm-l2-lineage-router/SKILL.md
Popularity
- Repository stars
- 51
- Repository forks
- 17
- Repository npm downloads, 7 days
- 48
- 7-day trend score
- 30
Stars and forks are for the whole GitHub repository, not this skill alone. Downloads are for the repository's npm package. Refreshed 17 Jun 2026, 4 months ago.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/shynlee04/hivemind-plugin/hm-l2-lineage-router)