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/Filip-Podstavec/claude-leverageWrote 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/filip-podstavec/claude-leverage/test)<a href="https://agentmods.dev/commands/filip-podstavec/claude-leverage/test"><img src="https://agentmods.dev/badge/commands/filip-podstavec/claude-leverage/test/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/filip-podstavec/claude-leverage/test"><img src="https://agentmods.dev/badge/commands/filip-podstavec/claude-leverage/test.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 | $0.00022 | $0.00575 |
| Opus 5.5 | $0.00009 | $0.00230 |
| Sonnet 5.5 | $0.00004 | $0.00115 |
| Haiku 4.5 | $0.00002 | $0.00057 |
Grade A, and why
test 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 — 51 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Context
Status: !git status --short
Recently changed files: !git diff --name-only HEAD~1 2>/dev/null || git diff --name-only
Your role
You are orchestrating a two-phase test workflow:
- Phase 1: a Sonnet-based subagent runs the tests and returns a structured report.
- Phase 2: you (the main session) apply approved fixes for any failures.
You do NOT run tests yourself. The subagent does that. Your job is to delegate, summarize, get user direction, and then make changes.
Workflow
-
Determine test scope:
- If
$ARGUMENTScontains paths or patterns, pass those to the subagent. - Otherwise let the subagent decide based on recent changes (its default behavior).
- If
-
Invoke the
test-runnersubagent with the scope. Wait for its structured report. -
If all tests passed: relay that to the user briefly and stop. Do not propose unsolicited improvements.
-
If tests failed: present the report's Summary and Failures sections to the user. Keep it concise - do not paraphrase the subagent's analysis, just relay it. If the subagent identified shared root causes (Patterns section), highlight that.
-
Ask for direction. Offer these options:
- Apply suggested fixes for all failures
- Apply fixes for specific failures (user picks)
- Skip fixes, the user will investigate
- Re-run a specific subset of tests (delegates back to the subagent)
Wait for explicit confirmation. Never apply fixes preemptively.
-
Apply approved fixes using your full toolset (Edit, Write, Bash). Use the subagent's "Suggested direction" as a starting point, not a prescription - you have the full reasoning context.
-
After fixes are applied, ask the user whether to re-run the affected tests. Do not auto-rerun.
Hard rules
- Do not run tests yourself in the main session. Always delegate to the subagent.
- Do not ask the subagent to "also fix the failures" - it is read-only by design and refusing is its core contract.
- Do not disable or skip failing tests as a workaround unless the user explicitly requests it.
- If the subagent reports a setup/configuration issue (missing deps, broken config), do not try to "work around" by guessing test commands. Address the setup issue first.
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 · 51 lines · 22 tokens per session scan A 7ec2f0d6cda0
test is a command published in the GitHub repository Filip-Podstavec/claude-leverage (68 stars, last pushed 2mo ago), licensed MIT. It adds 22 tokens to every session and 575 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 commands, from other repositories
prototype
You are building a proof-of-concept for the current Grainulator sprint. Read CLAUDE.md for sprint context and claims.json for existing research claims.
test
Generate a test scaffold for a file or function following the testing skill protocol — boundary, happy path, edge cases, error cases. Does not run; produces the file.
coder
Autonomous code generation with batch file operations.
tdd
Test-driven development with TodoWrite planning and comprehensive testing.
tester
Comprehensive testing with parallel execution capabilities.
test
Start Phase 5 — execute functional tests, log defects, report coverage, validate UX conformance.