delegation-audit — Claude Code skill: delegation audit
delegation-audit is a Claude Code skill that reviews which models handled each subagent type and how those assignments performed. It flags weak results, repeated failures, and possible model mismatches.
MIT ★ 1,198 updated 2026-10-23
/plugin marketplace add hoangsonww/Claude-Code-Agent-Monitor
/plugin install ccam-workflows@claude-code-agent-monitor-pluginsWhat delegation-audit does
For a session, the skill builds a model-by-subagent matrix and reports counts, success rates, and average durations by type. If you provide no session ID, it fetches the latest session and identifies it in the report.
The audit calls out low-yield types, repeated unsuccessful runs, and cases where a heavier model handled simple work. It suggests reassignment only when the session data supports it. The workflow uses the Agent Monitor API and needs the dashboard to be reachable; if it is not, the instructions say to run npm start from the repository root. It does not derive success rates when the effectiveness data does not provide them.
What it helps with
- map models to subagent types in a session
- spot subagent types with low success rates
- find repeated unsuccessful delegations
- review heavy-model assignments for simple tasks
Example prompts for the agent
- Audit delegation for session
SESSION_IDand show the model-to-subagent matrix. - Review the latest session and compare success rates across subagent types.
- Find repeated failed delegations in session
SESSION_IDand suggest evidence-backed model changes.
How to install delegation-audit
1Plugin marketplace
/plugin marketplace add hoangsonww/Claude-Code-Agent-Monitor/plugin install ccam-workflows@claude-code-agent-monitor-pluginsThis installs the whole ccam-workflows plugin with all its skills
2Skills CLI (Vercel)
npx skills add hoangsonww/Claude-Code-Agent-Monitor --skill delegation-audit -a claude-code3Manually: put the files in place
~/.claude/skills/delegation-audit/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/delegation-audit/
1Skills CLI (Vercel)
npx skills add hoangsonww/Claude-Code-Agent-Monitor --skill delegation-audit -a codex2Manually: put the files in place
.agents/skills/delegation-audit/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add hoangsonww/Claude-Code-Agent-Monitor --skill delegation-audit -a cursor2Manually: put the files in place
.cursor/skills/delegation-audit/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install hoangsonww/Claude-Code-Agent-Monitor plugins/ccam-workflows/skills/delegation-audit2Manually: put the files in place
.github/skills/delegation-audit/SKILL.md1Skills CLI (Vercel)
npx skills add hoangsonww/Claude-Code-Agent-Monitor --skill delegation-audit -a gemini-cli2Platform CLI
gemini skills install https://github.com/hoangsonww/Claude-Code-Agent-Monitor --path plugins/ccam-workflows/skills/delegation-auditManually: put the files in place
.devin/skills/delegation-audit/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about delegation-audit
How do I install delegation-audit in Claude Code?
Run /plugin marketplace add hoangsonww/Claude-Code-Agent-Monitor, then run /plugin install ccam-workflows@claude-code-agent-monitor-plugins.
What does delegation-audit do?
It maps models to subagent types for a session, reports success rates and durations, and flags weak or repeated delegations. It recommends model changes only when the session data supports them.
What happens if I omit the session ID?
The skill fetches the most recent session through the API and names the session it audits.
Source text
SKILL.md · MIT
---
name: delegation-audit
description: >
Audit model delegation and subagent effectiveness for a session — which
models handled which subagent types, per-type success rates and average
durations, and wasted delegations (heavy models on trivial work or types
that consistently fail) — using the Agent Monitor workflow intelligence API.
Use when reviewing how a session delegated work across models and subagents.
---
# Delegation Audit
Audit how a Claude Code session delegated work: model-to-subagent mapping and whether each delegation paid off.
## Input
The user provides: **$ARGUMENTS**
A session ID. If empty, fetch `GET /api/sessions?limit=1` and audit the most recent session, stating which one.
## Data Sources
| Endpoint | Returns |
|----------|---------|
| `GET /api/workflows/{sessionId}` | The `modelDelegation` dataset (which models are delegated which subagent types) and the `effectiveness` dataset (per-type completion/success rate, avg duration, task success) |
| `GET /api/agents` | Raw subagent records (`type`, `model`, `status`, `depth`, `parent`) to corroborate counts and statuses |
## Report Sections
### 1. Delegation Matrix
From `modelDelegation`: a model × subagent-type table of how many agents of each type each model ran.
| Model | explore | code-review | debugger | ... | Total |
|-------|---------|-------------|----------|-----|-------|
### 2. Effectiveness by Subagent Type
From `effectiveness`: per type, the success rate and average duration.
| Subagent type | Count | Success rate | Avg duration | Verdict |
|---------------|-------|--------------|--------------|---------|
Mark types below ~70% success as low-yield.
### 3. Wasted Delegations
Flag, with evidence:
- A heavy model (e.g. Opus) assigned to a simple/low-stakes subagent type that a cheaper model handled successfully elsewhere — candidate for rebalancing.
- Subagent types with low success rates (effort spent, task not completed).
- Duplicate delegations: the same type spawned repeatedly with poor success (retry churn).
### 4. Rebalancing Suggestions
Concrete model reassignments grounded in the matrix and effectiveness data. State the type, the model used, the success rate, and the suggested model — only where the data supports it.
## Output
- Markdown tables for the matrix and effectiveness.
- Success rates as percentages; durations in human units (e.g. `1m 12s`).
- Use ▲/▼ when comparing a type's success rate against the session-wide average.
- Cite only numbers returned by the API; do not infer success rates that the `effectiveness` dataset does not provide.
- If the dashboard is unreachable, tell the user to start it with `npm start` from the repo root.
© Son Nguyen · MIT · hoangsonww/Claude-Code-Agent-Monitor