Borrowing it
Nothing to install: this file belongs to Stankye/profiler-mcp. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/Stankye/profiler-mcp/main/.claude/skills/profiler-selftest/SKILL.mdgit clone --depth 1 https://github.com/Stankye/profiler-mcpWrote 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/stankye/profiler-mcp/profiler-selftest)<a href="https://agentmods.dev/skills/stankye/profiler-mcp/profiler-selftest"><img src="https://agentmods.dev/badge/skills/stankye/profiler-mcp/profiler-selftest/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/stankye/profiler-mcp/profiler-selftest"><img src="https://agentmods.dev/badge/skills/stankye/profiler-mcp/profiler-selftest.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 | $0.00068 | $0.00681 |
| Opus 5 | $0.00034 | $0.00341 |
| Sonnet 5 | $0.00014 | $0.00136 |
| Haiku 4.5 | $0.00007 | $0.00068 |
Grade A, and why
profiler-selftest 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 9d 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 — 50 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Profiler self-test
Everything is testable on any machine: the mock CLIs (src/profiler_mcp/mocks/) stand in
for the real profilers with byte-realistic output (see docs/research/captures-*.md).
Fast path
uv run pytest # layers 1-4: unit, in-memory MCP, stdio smoke, ground truth
uv run ruff check src tests
Both must be green before and after any change. If pytest is green but you changed mock
output shapes, also re-check parser fixtures — fixtures are verbatim vendor output and
must NEVER be edited to make a parser pass (fix the parser; fixtures only change when new
verbatim captures are added to docs/research/).
The layers (docs/SELF_TESTING.md has detail)
| Layer | Where | Catches |
|---|---|---|
| 1 unit | tests/test_core_*, parser tests |
parsing/logic bugs |
| 2 in-memory MCP | tests/test_vtune_*, test_uprof_* |
tool schema/wiring bugs |
| 3 stdio smoke | tests/test_stdio_smoke.py |
transport/entry-point bugs |
| 4 ground truth | tests/test_ground_truth.py |
end-to-end attribution: profile the compiled examples/workloads/hotspots binary, assert hot_compute dominates per the sidecar model; seeded regression must fail *_compare |
| 5 agentic | .claude/workflows/self-test.js (Workflow tool) |
LLM-facing ergonomics: unclear errors, oversized outputs, schema surprises |
Layer 5: agentic self-test
Run via the Workflow tool with {scriptPath: ".claude/workflows/self-test.js"}. It runs
layers 1–4, then fans out probe agents that drive each server through MCP as a model
would (happy paths, error paths, fault-injected mocks via MOCK_FAIL), and adversarially
verifies every reported failure before surfacing it. Findings that survive verification
are real bugs or real ergonomics problems — feed them to
.claude/workflows/optimize-project.js.
Debugging failures
- One mock's behavior is off → run it directly:
uv run python src/profiler_mcp/mocks/vtune_mock.py -collect hotspots -result-dir /tmp/r -- ./examples/workloads/hotspots --seconds 0.3 - MCP layer failure → reproduce with the in-memory client (
create_connected_server_and_client_session) in a scratch script before touching test code. - Fault-injection paths: set
MOCK_FAIL=collect-perm|collect-hang|report-empty|bad-exit. - Determinism: mocks honor
MOCK_FIXED_TIME=<unix-sec>for byte-stable timestamps.
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.
- 9d ago First seen · 50 lines · 68 tokens per session scan A 728d64929160
profiler-selftest is a skill published in the GitHub repository Stankye/profiler-mcp (0 stars, last pushed 1mo ago), licensed MIT. It adds 68 tokens to every session and 681 once invoked, about $0.0003 per session on Opus 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 skills, from other repositories
axiom-testing
Use when writing ANY test, debugging flaky tests, making tests faster, or choosing Swift Testing vs XCTest. Covers unit tests, UI tests, async testing, test architecture.
designing-tests
Designs and implements testing strategies for any codebase. Use when adding tests, improving coverage, setting up testing infrastructure, debugging test failures, or when asked about unit tests, integration tests, or E2E testing.
test-automation
Execute Vitest and Playwright test suites with result collection and failure analysis.
testing-blocks
Use this when you have made AEM Edge Delivery Services code changes to blocks, scripts, or styles and need to validate them before opening a pull request. Covers unit testing for utilities and logic, browser testing with Playwright, linting, and guidance on what to test and how.
prd-auto-test-loop
A testing workflow driven by a product requirements document (PRD), which describes what a software version should do. It turns acceptance criteria into unit, integration, and end-to-end tests, then records the plan and results.
test-automation-expert
Comprehensive test automation specialist covering unit, integration, and E2E testing strategies. Expert in Jest, Vitest, Playwright, Cypress, pytest, and modern testing frameworks. Guides test pyramid design, coverage optimization, flaky test detection, and CI/CD integration. Activate on 'test strategy', 'unit tests'…