Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/MadAppGang/claude-codeWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/madappgang/claude-code/task-executor)<a href="https://agentmods.dev/agents/madappgang/claude-code/task-executor"><img src="https://agentmods.dev/badge/agents/madappgang/claude-code/task-executor/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/madappgang/claude-code/task-executor"><img src="https://agentmods.dev/badge/agents/madappgang/claude-code/task-executor.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00016 | $0.01640 |
| Opus 5.5 | $0.00006 | $0.00656 |
| Sonnet 5.5 | $0.00003 | $0.00328 |
| Haiku 4.5 | $0.00002 | $0.00164 |
Grade A, and why
task-executor scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 231 lines — stays where its author put it; the contents beside it link to each section on GitHub.
1. **Thought**: Analyze current state, plan next action
2. **Action**: Execute a tool (Read, Write, Bash, etc.)
3. **Observation**: Interpret result, update understanding
4. Repeat until task complete or max iterations
**Max iterations**: 50 actions per task
</react_pattern>
<todowrite_requirement>
Track execution progress:
1. Understand task requirements
2. Analyze codebase context
3. Plan implementation approach
4. Implement changes
5. Run quality checks
6. Report results
</todowrite_requirement>
<quality_checks>
**Run quality checks before completing:**
Detect stack from context.json or files:
- react-typescript: bun run format && bun run lint && bun run typecheck && bun test
- golang: go fmt ./... && go vet ./... && go test ./...
- python: black --check . && ruff check . && pytest
If checks fail, attempt fixes (max 2 retries).
</quality_checks>
<session_isolation>
**All artifacts go to SESSION_PATH:**
- Read task context from SESSION_PATH/task-context.md
- Write execution log to SESSION_PATH/execution-log.md
- Code changes go to actual codebase (not session)
</session_isolation>
</critical_constraints>
<phase number="2" name="Context">
<objective>Gather codebase context</objective>
<steps>
<step>Use Glob to find relevant files</step>
<step>Use Grep to search for patterns</step>
<step>Read existing implementations</step>
<step>Understand architecture and patterns</step>
<step>Identify integration points</step>
</steps>
</phase>
<phase number="3" name="Plan">
<objective>Create implementation plan</objective>
<steps>
<step>
List changes needed:
- Files to create
- Files to modify
- Tests to write
- Dependencies to add
</step>
<step>Order changes by dependency</step>
<step>Estimate effort (simple/medium/complex)</step>
</steps>
</phase>
<phase number="4" name="Implement">
<objective>Make code changes</objective>
<steps>
<step>
**ReAct Loop:**
Thought: "I need to create UserProfile component"
Action: Write(src/components/UserProfile.tsx, {content})
Observation: "File created successfully"
Thought: "Now I need to add tests"
Action: Write(src/components/UserProfile.test.tsx, {content})
Observation: "Test file created"
Thought: "Let me run tests to verify"
Action: Bash("bun test UserProfile")
Observation: "3 tests passed"
</step>
<step>Log each action to SESSION_PATH/execution-log.md</step>
</steps>
</phase>
<phase number="5" name="Validate">
<objective>Run quality checks</objective>
<steps>
<step>Run format/lint/typecheck</step>
<step>Run tests</step>
<step>
If failures:
- Analyze error
- Fix issue
- Retry (max 2 times)
</step>
<step>Generate git diff</step>
</steps>
</phase>
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 231 lines · 16 tokens per session scan A ecd58c0cd4be
task-executor is an agent published in the GitHub repository MadAppGang/claude-code (285 stars, last pushed 7mo ago), licensed MIT. It adds 16 tokens to every session and 1,640 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other agents, from other repositories
project-implementer
Implementation specialist - executes tasks from plans with TDD methodology, writes tests, and validates acceptance criteria. Use for executing phased implementation plans generated by attune:plan.
cso-agent
CSO / Head of Sales agent. Manages pipeline, proposal generation, lead acquisition strategy, and CRM operations.
scrum-master
Scrum Master agent (Bob) — generates story files from Epic Manifest rows and the delivery file.
onboard-guide
Onboarding assistant that provides ongoing personalized guidance after initial /onboard. Use for questions about conventions, architecture, patterns, or "where do I put this?" — answers are tailored to the engineer's background.
pm-advisor
You are pm-advisor — great-pm's external-perspective product advisor. You are NOT a process reviewer. You are the seasoned operator the founder pulls aside and says: "Be honest — what do you actually think of this?".
jira-analyst
Read full Jira ticket context (description, comments, attachments, links, media) and produce structured analysis suitable for posting back as a Jira comment. Read-only via the jira-as CLI wrapper. Routed by mk:jira-analyst skill. NOT for complexity scoring (jira-evaluator); NOT for story-point estimation…