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 mystilleef/spae-framework --skill cleangit clone --depth 1 https://github.com/mystilleef/spae-frameworkWrote 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/mystilleef/spae-framework/clean)<a href="https://agentmods.dev/skills/mystilleef/spae-framework/clean"><img src="https://agentmods.dev/badge/skills/mystilleef/spae-framework/clean/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/mystilleef/spae-framework/clean"><img src="https://agentmods.dev/badge/skills/mystilleef/spae-framework/clean.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 | $0.00013 | $0.00852 |
| Opus 5.5 | $0.00005 | $0.00341 |
| Sonnet 5.5 | $0.00003 | $0.00170 |
| Haiku 4.5 | $0.00001 | $0.00085 |
Grade C, and why
clean scanned grade C with 1 finding 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 3d 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.
Hidden instructionshighPrompt injection
Directives inside HTML comments, invisible characters or bidirectional overrides are read by the model and not by the person reviewing the file.
<!-- prettier-ignore-start --> How it starts
The opening of the file, as written. The whole thing — 114 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Clean
When to use
- The user wants to clean smelly code autonomously using the
purifyandrefactoragents.
Goal
- Run purification and refactoring
subagentssequentially to clean target code.
Input
- Accept optional arguments (file or folder paths).
Workflow
- Purification Loop:
- Immediately use the
subagenttool to spawn apurifysubagent(pass optional arguments). - Await purification result.
- If
subagentreturnsNo Changesstatus, exit loop and proceed. - On
Failedstatus or any process failure, halt immediately and surface the error. - Loop back to spawn another
purifysubagentif the previous pass returnsComplete.
- Immediately use the
- Refactoring Loop:
- Immediately use the
subagenttool to spawn arefactorsubagent(pass optional arguments). - Await refactoring result.
- If
subagentreturnsNo Changesstatus, exit loop and finish. - On
Failedstatus or any process failure, halt immediately and surface the error. - Loop back to spawn another
refactorsubagentif the previous pass returnsComplete.
- Immediately use the
Directives
- Always use the subagent tool for subagent invocation.
- Pass the optional argument verbatim to both subagents; never read or expand file-path arguments.
- Rely on
subagentstatus blocks to direct the loop. - Halt immediately on subagent
Failedstatus or any process failure (crash, timeout,AgentError). - List of agents permitted to invoke:
purifyrefactor
Constraints
- Restrict your activities to:
- Using the
subagenttool to spawn agents. - Using agent results to direct the loop or advance the workflow.
- Using the
- Perform orchestration only; never edit codebase files, tests, or documentation.
- Operate strictly in read-only mode.
- Perform purification to completion first, before starting refactoring.
- Never interleave purification with refactoring.
- Never run
subagentsin parallel or concurrently. - Never activate skills.
- Never invoke agents outside the permitted list.
- Never prompt the user for decisions mid-run; let blockers halt execution.
- Never perform activities beyond subagent invocation, state tracking, and reporting.
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.
- 3d ago Changed b3280574e70b
- 1mo ago First seen · 114 lines · 13 tokens per session scan C d2ef839aaecd
clean is a skill published in the GitHub repository mystilleef/spae-framework (2 stars, last pushed 3d ago), licensed MIT. It adds 13 tokens to every session and 852 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it C with 1 finding (hidden instructions). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
ahk-review
Preview a code review against ticket/objective alignment, with deep semantic (name-vs-behavior) analysis. No tasks created, no harness tracking.
red-team-assessment
Run an evidence-backed adversarial security assessment through the local llm-cli-gateway stdio MCP server. Use for code, architecture, configuration, data-flow, or supply-chain security review across the full Claude, Codex, Gemini, Grok, Mistral, Devin, and Cursor roster.
multi-llm-consensus
Obtain complete, evidence-backed agreement from a required roster of llm-cli-gateway providers. Use for high-stakes generation, conflict resolution, or final quality gates that require unconditional approval from every required reviewer.
agent-codex-gate
Pattern for spawning subagents whose work must receive an evidence-backed, unconditional Codex review through the local llm-cli-gateway stdio MCP server. Use when parallel implementation tasks need an independent Codex quality gate.
codex-review-gate
Submit completed work to Codex through the local llm-cli-gateway stdio MCP server and continue until an evidence-backed, unconditional approval. Use after implementation, before merging, or whenever a process requires Codex sign-off.
python-engineering-guidelines
General Python engineering standards for writing, reviewing, and refactoring maintainable code. Use when Codex works on Python code, especially in medium/large projects, runtime adapters, APIs, SDK integrations, async services, stateful systems, or refactors where readability, explicit typing, module boundaries, side…