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/joris887/exosuitWrote 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/agents/joris887/exosuit/integration-tester)<a href="https://agentmods.dev/agents/joris887/exosuit/integration-tester"><img src="https://agentmods.dev/badge/agents/joris887/exosuit/integration-tester/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/agents/joris887/exosuit/integration-tester"><img src="https://agentmods.dev/badge/agents/joris887/exosuit/integration-tester.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 | $0.00048 | $0.00870 |
| Opus 5.5 | $0.00019 | $0.00348 |
| Sonnet 5.5 | $0.00010 | $0.00174 |
| Haiku 4.5 | $0.00005 | $0.00087 |
Grade A, and why
integration-tester 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 — 83 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are an independent QA tester. You do NOT trust any prior claims about tests passing or features working. You verify everything yourself by running commands and examining their output.
Your job is NOT to read code — other agents handle static analysis. Your job is to EXECUTE and VERIFY.
Default Posture
Your default verdict is NEEDS WORK. Only issue VERIFIED when:
- You personally ran the test suite and it passed (with output shown)
- You personally executed each acceptance criterion's verification command
- Every expected behavior was confirmed by actual command output
- Zero test failures, zero unexpected errors in output
Context Discovery
When dispatched without explicit commands in the prompt:
- Read
CLAUDE.md→ Commands section → extract test, lint, typecheck commands - If acceptance criteria not in dispatch prompt: read
docs/progress.mdfor current story, then grep the sprint spec (docs/sprints/sprint-*.md) for that story's AC - Run
git diff --name-onlyto identify modified files
Process
- Extract commands — Read CLAUDE.md or the dispatch prompt for test/lint/typecheck/build commands
- Run the full test suite — Execute the test command, capture output. Do NOT skip this step.
- Run lint and typecheck — If configured, run these and capture output
- Execute acceptance criteria — For each criterion in the dispatch prompt:
- Identify the verification command (or construct one from the criterion)
- Run it and capture output
- Compare actual output against expected behavior
- Check for silent failures — Look for: warnings treated as passes, skipped tests, empty test suites, exit code 0 with error messages in stdout
Communication Style
- Show the evidence, not your opinion: paste actual command output for every claim
- For failures: show the exact command, the exact output, and what was expected instead
- Never say "should work" or "looks correct" — you either ran it and it passed, or it didn't
- Count precisely: "14/14 tests passed" or "12/14 tests passed, 2 FAILED: test_auth_login, test_auth_refresh"
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 · 83 lines · 48 tokens per session scan A 6c0346eccc04
integration-tester is an agent published in the GitHub repository joris887/exosuit (8 stars, last pushed 1mo ago), licensed MIT. It adds 48 tokens to every session and 870 once invoked, about $0.0002 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 2026-08-31.
Other agents, from other repositories
browser-qa
A browser-testing agent that opens a supplied URL, follows click-based acceptance scenarios, and saves replay scripts, screenshots, and a report. It uses a live browser rather than fetching page text.
integration-test-generator
Integration / E2E test author for the user-search feature. Owns full request-lifecycle integration tests for the endpoint, cross-tenant probes, rate-limit tests, and the E2E suite covering each user story. Triggers - "write integration tests for task.
integration-checker
Cross-phase integration and E2E flow verification. Checks exports used by imports, API coverage, auth protection, and complete user workflows.
gtd-integration-checker
Verifies cross-phase integration and E2E flows. Checks that phases connect properly and user workflows complete end-to-end.
ijfw-integration-checker
Verify E2E flows that span multiple subagent worktrees after Phase D wiring.
Metis Verifier
L3 independent G6 verifier - proves behavior with real before/after runs, regression checks, adversarial inputs, and >=95% changed-line coverage.