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 vincentkoc/dotskills --skill crabpot-perf-metricsgit clone --depth 1 https://github.com/vincentkoc/dotskillsWrote 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/vincentkoc/dotskills/crabpot-perf-metrics)<a href="https://agentmods.dev/skills/vincentkoc/dotskills/crabpot-perf-metrics"><img src="https://agentmods.dev/badge/skills/vincentkoc/dotskills/crabpot-perf-metrics/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/vincentkoc/dotskills/crabpot-perf-metrics"><img src="https://agentmods.dev/badge/skills/vincentkoc/dotskills/crabpot-perf-metrics.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 | $0.00040 | $0.01055 |
| Opus 5 | $0.00020 | $0.00528 |
| Sonnet 5 | $0.00008 | $0.00211 |
| Haiku 4.5 | $0.00004 | $0.00105 |
Grade A, and why
crabpot-perf-metrics 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 11d 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 — 73 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Purpose
Turn Crabpot/OpenClaw perf artifacts into a clear read: what changed, what is real signal, what is stale/noisy, and what action is worth taking.
When to use
- Comparing
openclaw/crabpotbranches such asmain,crab-beta, orcrab-development. - Reading
reports/crabpot-dashboard-data.json,reports/crabpot-import-loop-profile.json,reports/crabpot-runtime-profile.json, README dashboard metrics, or GitHub Actions report artifacts. - The user asks about perf, runtime, CPU, RSS, import loop, OpenClaw import/activate, dashboard deltas, or whether a Crabpot performance change is meaningful.
Inputs
- Target repo or local checkout, usually
openclaw/crabpot. - Branches, refs, PRs, or artifact URLs to compare.
- Optional target OpenClaw labels/SHAs if the comparison crosses
openclaw@latest, beta, oropenclaw/openclaw@main.
Workflow
- Establish freshness before interpreting numbers.
- Resolve branch heads and report timestamps.
- Compare dashboard JSON with raw profile JSON on the same branch.
- If one branch is stale, say so first and avoid strong conclusions.
- Prefer raw JSON over README snippets when they disagree.
- Separate the metric families.
import-loop-profile: cheap cold capture loop against a tiny fixture. Good for harness/import drift, weak for production runtime claims.runtime-profile: macro command timings across Crabpot analysis commands. Better trend signal for real suite cost.OpenClaw lifecycle: import plus activate phases. IfopenClawLifecycleCountis0, the branch dashboard does not include this signal.
- Interpret import-loop metrics conservatively.
- Use
p50WallMs,p95WallMs,p50PluginWallDeltaMs,p95PluginWallDeltaMs,maxPluginPeakRssDeltaMb, andmaxPluginCpuDeltaMsEstimate. - Treat 3-run p95 as "worst of three", not a stable tail latency estimate.
- A small wall-time bump with flat plugin RSS/CPU usually means jitter or module-resolution overhead, not a memory/CPU regression.
- Baseline-adjusted plugin deltas are more useful than raw wall/RSS/CPU values.
- Use
- Interpret runtime profile metrics as the stronger trend.
- Use
summary.p50WallMs,summary.p95WallMs,summary.maxPeakRssMb,summary.maxCpuMsEstimate, andsummary.maxHarnessHeapDeltaMb. - Inspect per-command medians for the source of change:
fixture-inspection,compat-report-registry,contract-capture,synthetic-probe-plan,cold-import-readiness,workspace-plan,platform-probes, andimport-loop-profile. - Broadly uniform slowdowns across target-aware commands usually point at target OpenClaw surface parsing or registry work, not one plugin fixture.
- If RSS is flat and CPU moves only slightly, call it a wall-clock/runtime overhead change, not a memory leak.
- Use
- Account for surface changes.
- Different fixture counts, entrypoint counts, OpenClaw labels, or commit SHAs make the comparison partly apples-to-oranges.
- New fixtures can raise issue/probe counts while leaving perf healthy.
- OpenClaw
mainvs npmlatestcan improve compatibility while adding modest analysis overhead.
- Give the user a direct read.
- Start with the conclusion.
- Prefer bullets and a short summary over tables unless the user asks for a table.
- Include exact deltas and classify them:
real regression,modest slowdown,noise,stale data, ormissing signal. - Recommend one next action: refresh dashboard, add a dedicated lifecycle lane, investigate a specific command, or ignore as noise.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 11d ago First seen · 73 lines · 40 tokens per session scan A 0781a6d101e5
crabpot-perf-metrics is a skill published in the GitHub repository vincentkoc/dotskills (102 stars, last pushed 2d ago), licensed MIT. It adds 40 tokens to every session and 1,055 once invoked, about $0.0002 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 2026-08-30.
Other skills, from other repositories
taiyi-plan
A TaiyiForge planning step that turns a README, product requirements document, or technical plan into a list of separate project changes. Each change includes a suggested work profile, dependencies, and priority.
taiyi-integration
A project workflow skill for closing a TaiyiForge change and recording it in a CHANGELOG.md file. It checks review results, tests, and the state of the working tree before archiving the change.
taiyi-requirement
A requirements-analysis process that turns a proposed change into a REQUIREMENT.md document. It records user needs, acceptance checks, terminology, triggers, and dependencies, with different levels of detail for different project sizes.
security-compliance
Guides security professionals in implementing defense-in-depth security architectures, achieving compliance with industry frameworks (SOC2, ISO27001, GDPR, HIPAA), conducting threat modeling and risk assessments, managing security operations and incident response, and embedding security throughout the SDLC.
taiyi-change
A TaiyiForge workflow stage for writing a change proposal in CHANGE.md, with checks for scope, risks, success criteria, and related project documents.
taiyi-test
A project workflow skill for verifying an implementation and producing a TEST.md record. TDD means writing a failing test, implementing the change, and then making the test pass; this skill checks that process and other regression cases.