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/MarioEpkOne/PipelineIQWrote 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/commands/marioepkone/pipelineiq/impl)<a href="https://agentmods.dev/commands/marioepkone/pipelineiq/impl"><img src="https://agentmods.dev/badge/commands/marioepkone/pipelineiq/impl/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/commands/marioepkone/pipelineiq/impl"><img src="https://agentmods.dev/badge/commands/marioepkone/pipelineiq/impl.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 | $0.00000 | $0.03086 |
| Opus 5.5 | $0.00000 | $0.01234 |
| Sonnet 5.5 | $0.00000 | $0.00617 |
| Haiku 4.5 | $0.00000 | $0.00309 |
Grade A, and why
impl 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 2mo 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 — 160 lines — stays where its author put it; the contents beside it link to each section on GitHub.
EXECUTE PLAN — reads an existing implementation plan and runs all steps autonomously, verifies each step, writes a working log, and prompts to commit.
The user's request is: $ARGUMENTS
Can be run from the main repo or from inside the worktree directory. If the plan has a
**Worktree:**header, the skill automatically resolves and uses that worktree as its working root.
Process
Phase 1 — Find the Implementation Plan
Search Implementation Plans/ (relative to the current working directory) for a file whose name contains the argument (case-insensitive).
- If multiple matches: list them and ask the user to pick one.
- If no match: tell the user to run
/impl-planfirst and exit. - If no argument was provided: list all impl plans in
Implementation Plans/and ask the user which one to execute.
Read the full implementation plan.
Phase 2 — Verify Worktree
Read the **Worktree:** field from the plan header.
- No worktree field in the plan: proceed from the current working directory.
- Worktree field present, CWD already ends with that slug: proceed normally.
- Worktree field present, CWD is the main repo: resolve the absolute path of the worktree (the field value is relative to the main repo root, e.g.
.claude/worktrees/my-feature/). Check whether<worktree_abs_path>/Implementation Plans/<same-filename>exists — if it does, re-read the plan from there (the worktree copy may differ from the main repo copy, so prefer it). Whether or not a worktree copy exists, use<worktree_abs_path>as the working root for all subsequent file reads, writes, and relative path references in this session. Do not exit — continue with the next phase.
Phase 3 — Read Reading List (and optionally learnings.md)
Read every file listed in the plan's Reading list section, in order.
Do not read any other files unless a specific step explicitly requires it and explains why.
Optional — learnings.md: If learnings.md exists in the project root and the task involves an area with known gotchas, skim it for relevant entries. Treat them as risks to watch for, not hard rules.
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.
- 2mo ago First seen · 160 lines · 0 tokens per session scan A aad740dfd0c3
impl is a command published in the GitHub repository MarioEpkOne/PipelineIQ (2 stars, last pushed 4mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 3,086 tokens. 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 commands, from other repositories
review-plan
Critically review the current plan/approach before implementing — assumptions, gaps, blast radius, deploy safety.
explain-dev
Write a WhatsApp-ready, non-technical message explaining a development to the client.
feature
Run the full feature pipeline — explore → plan → implement → review → test — for the change described in the argument.
triage
Triage a bug report (PDF / markdown / text) — extract the issues, prioritize, then run the fix pipeline per issue.
fix
Diagnose and fix a bug — reproduce → locate → root-cause → fix → test.
review
Review the current diff for bugs, security issues, and regressions.