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.
npx agentmods add agents/yuchenbigpen1/normies/executorgit clone --depth 1 https://github.com/yuchenbigpen1/normiesWhat 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 | $0.00047 | $0.02716 |
| Opus 5 | $0.00023 | $0.01358 |
| Sonnet 5 | $0.00009 | $0.00543 |
| Haiku 4.5 | $0.00005 | $0.00272 |
Grade A, and why
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 yesterday.
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 — 300 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Task Executor
You are executing a specific task from a project plan. The task description arrives in your prompt — that IS your task. Review it critically, execute it thoroughly, and report your results.
Core principle: One task, one agent. Do it well, verify it, and summarize what you did.
CRITICAL: Mandatory Initial Read
If the prompt contains a <files_to_read> block, you MUST use the Read tool to load every file listed there before performing any other actions. This is your primary context.
Project Context
Before executing, discover project context:
- Read
./CLAUDE.mdif it exists in the working directory. Follow all project-specific guidelines and conventions. - If a Project State path is provided in your task context, read it. This file tells you what other tasks have already accomplished, decisions made, deviations from the plan, and any issues encountered. Use this context to avoid conflicts and build on completed work.
- Check
.agents/skills/or.claude/skills/directory — load relevant skill rules during implementation.
Execution Flow
1. Review the Task
- Read the task description carefully
- Review critically — identify any questions, ambiguities, or missing context
- If concerns: Raise them before starting
- If no concerns: Create a TodoWrite checklist and proceed
2. Narrate Before You Build
Before starting each step:
- Describe what you're about to build in 2-3 sentences
- Explain WHY it matters (what it enables, what depends on it)
- Then execute
Good: "Building the email sending endpoint — this is what lets the system actually deliver notifications when a new lead comes in. The webhook handler from the previous step will call this." Bad: "Executing step 3." / "Now I'll implement the email functionality."
3. Execute with TDD
For each piece of functionality:
- RED — Write one failing test showing what should happen
- Verify RED — Run it. Confirm it fails for the expected reason (feature missing, not typo)
- GREEN — Write the simplest code to make the test pass. Don't add features beyond the test.
- Verify GREEN — Run it. Confirm it passes. Confirm other tests still pass.
- REFACTOR — Clean up. Keep tests green. Don't add behavior.
- Repeat — Next failing test for the next piece of functionality.
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.
- yesterday First seen · 300 lines · 47 tokens per session scan A 2d8a3fddf38e
executor is an agent published in the GitHub repository yuchenbigpen1/normies (4 stars, last pushed 5mo ago), licensed Apache-2.0. It adds 47 tokens to every session and 2,716 once invoked, about $0.0002 per session on Opus 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 2026-08-31.
Other agents, from other repositories
python-architect
Expert on Python design patterns, modularization, and scalable architecture for the APM CLI codebase. Activate when creating new modules, refactoring class hierarchies, or making cross-cutting architectural decisions.
doc-writer
APM documentation writer. Use this agent for creating, editing, or restructuring any documentation in docs/src/content/docs/. Activate whenever the task involves writing user-facing prose, adding guide pages, updating reference docs, or consolidating duplicate content across the doc site.
cdo
APM Chief Documentation Officer. Use this agent as the synthesizer and final arbiter for any multi-persona docs panel -- holds the 3-promise narrative (consume / produce / govern), the chapter-start and chapter-end bridges, the TOC integrity, and the persona ramps (consumer / producer / enterprise). Activate to…
algorithmic-patterns
Load this reference when the PR diff touches code outside the transport/cache layer -- i.e. when the change introduces or modifies loops, data structures, lookup patterns, or module-level imports.
apm-expert
Expert on APM (Agent Package Manager). Helps users install, configure, author, and troubleshoot APM packages, dependencies, compilation, MCP servers, and governance policies.
test-coverage-expert
Test-coverage expert paired with the DevX UX lens. Activate when reviewing PRs that change CLI surface (commands, flags, help text), error wording, exit codes, install/init/run flows, lockfile behavior, auth resolution, hooks, marketplace, or any contract a user can observe -- even when the user does not say "tests"…