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 skills add code-yeongyu/senpi --skill tmux-manual-qagit clone --depth 1 https://github.com/code-yeongyu/senpiWrote 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/skills/code-yeongyu/senpi/tmux-manual-qa)<a href="https://agentmods.dev/skills/code-yeongyu/senpi/tmux-manual-qa"><img src="https://agentmods.dev/badge/skills/code-yeongyu/senpi/tmux-manual-qa/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/skills/code-yeongyu/senpi/tmux-manual-qa"><img src="https://agentmods.dev/badge/skills/code-yeongyu/senpi/tmux-manual-qa.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00066 | $0.01545 |
| Opus 5.5 | $0.00026 | $0.00618 |
| Sonnet 5.5 | $0.00013 | $0.00309 |
| Haiku 4.5 | $0.00007 | $0.00154 |
Grade A, and why
tmux-manual-qa 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 — 129 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Tmux Manual QA Worker
You are executing ONE manual QA feature from features.json that drives the real ./pi-test.sh CLI inside a tmux session, captures scrollback, and asserts deterministic pass/fail markers.
Context you MUST read before starting
- Feature spec:
features.json— your assigned feature. - Mission document:
mission.md. - Validation contract: the
fulfillsIDs for your feature invalidation-contract.md. - Architecture:
.factory/library/architecture.md. - User testing surface:
.factory/library/user-testing.md— especially the Manual tmux TUI section. - Mission AGENTS.md: boundaries and git safety rules.
Hard rules
- Real LLM calls are allowed in this skill (manual QA only). The user's
~/.pi/agent/auth.jsonis presumed configured. Do NOT touch that file. - Capture to
local-ignore/— never commit QA evidence. Thelocal-ignore/directory is gitignored. - Clean up test fixtures. If you create a temporary
.pi/settings.jsonfor a scenario, delete it after capture so subsequent tests start from a clean slate. - Deterministic evidence: every manual feature includes a
rg -c "SYSTEM DIRECTIVE: SENPI"count check. Always save the count to a.countfile alongside the.logfile. The count is the canonical pass/fail marker, not the visual scrollback. - No src/ changes: you are verifying only. If you find a bug, return to orchestrator with details and do NOT fix it yourself — a coding-agent-extension-worker will handle the fix in a follow-up feature.
Prerequisites
Before running any scenario, confirm:
./pi-test.shis executable and runs (checkls -la pi-test.sh).npm run buildhas been run at least once after the continuation feature was merged (checkpackages/coding-agent/dist/cli.jsexists and contains the continuation code).tmuxis installed (command -v tmux).rgis installed (command -v rg).local-ignore/directory exists at repo root (create if needed).
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 · 129 lines · 66 tokens per session scan A 847dbefaeb22
tmux-manual-qa is a skill published in the GitHub repository code-yeongyu/senpi (488 stars, last pushed today), licensed MIT. It adds 66 tokens to every session and 1,545 once invoked, about $0.0003 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 skills, from other repositories
llm-as-judge-evaluation
Evaluate LLM outputs using frontier models as judges. Use for pairwise model comparison, quality scoring with custom rubrics, and automated evaluation pipelines. Covers position bias mitigation, statistical significance, and generating preference data for DPO/RLHF.
transformers
This skill should be used when working with pre-trained transformer models for natural language processing, computer vision, audio, or multimodal tasks. Use for text generation, classification, question answering, translation, summarization, image classification, object detection, speech recognition, and fine-tuning…
reproduce
Reproduces a paper's result, a claim, an artifact or a previous run with the target and success criterion frozen first, the canonical code path run before any substitute, exact inputs, environment, seeds and commands captured, and a verdict from a fixed set, reproduced, reproduced within tolerance, partially…
acceptance-checks
Turns the acceptance criteria a task states into executable checks and runs them against the finished outputs before delivering, rather than trusting that a correct-looking analysis satisfies them; rehearses the hidden condition when the grader scores unseen instances from a described family; treats a stated formula…
submitted-code
Writes code that survives being re-executed by someone else on inputs you never saw, targeting the grader's or collaborator's runtime rather than your own, with one self-contained entry point using only the libraries the specification names, no reliance on packages you installed or files you cached during development…
pi-package-sandbox-test
Use when verifying that published Pi packages install and load for normal users, including pi.dev package compatibility checks, release validation, Daytona sandbox tests, and package installability audits.