Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/MichelKerkmeester/skilled-agent-harness_spec-driven-loopsnpx agentmods add skills/michelkerkmeester/skilled-agent-harness_spec-driven-loops/attachWrote 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/michelkerkmeester/skilled-agent-harness_spec-driven-loops/attach)<a href="https://agentmods.dev/skills/michelkerkmeester/skilled-agent-harness_spec-driven-loops/attach"><img src="https://agentmods.dev/badge/skills/michelkerkmeester/skilled-agent-harness_spec-driven-loops/attach/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/michelkerkmeester/skilled-agent-harness_spec-driven-loops/attach"><img src="https://agentmods.dev/badge/skills/michelkerkmeester/skilled-agent-harness_spec-driven-loops/attach.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 | $0.00041 | $0.02091 |
| Opus 5.5 | $0.00016 | $0.00836 |
| Sonnet 5.5 | $0.00008 | $0.00418 |
| Haiku 4.5 | $0.00004 | $0.00209 |
Grade A, and why
cli-classifier 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 16d 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 — 163 lines — stays where its author put it; the contents beside it link to each section on GitHub.
cli-classifier - Classifier Transport Hub
One public skill identity for classifier models that return typed judgments: the hosted Jev service and the Deem model served on this machine. This hub holds no packet-local logic. It routes by workflowMode through mode-registry.json and uses hub-router.json for router policy, signals and outcomes.
1. WHEN TO USE
Use this skill when a request needs a typed judgment from a classifier or needs to know whether one is usable.
| Mode | Packet | Kind | Use it for | Tool posture |
|---|---|---|---|---|
cli-jev |
cli-usage |
transport | A hosted Jev judgment through the jev CLI or its MCP surface: a probability, one option key, an ordered score position or a batch of keyed answers |
Read and Bash only, mutates nothing |
cli-deem |
cli-deem |
transport | A Deem availability check, a probability, one option key, an ordered level or a batch of keyed answers from the local Deem server | Read and Bash only, mutates nothing |
A caller names its backend. The two transports answer the same judgment types and never fail over to each other silently.
When NOT to Use
- Work that needs an edit, a review or a multi-step process. Dispatch the judgment to a transport and pair it with a workflow mode.
- Running
jev-mcpfrom a shell. Its stdio carries MCP frames, so it belongs in an MCP host configuration. - Starting, stopping or updating the Deem server. That is the operator's step with
deem-ctl.
2. SMART ROUTING
Routing is registry-driven. mode-registry.json lists every packet. hub-router.json decides whether the result is one transport, both transports in order or a defer.
Compiled routing (default-on, flag-gated, additive). Resolve the mode via the compiled router contract first:
node .skilled/bin/compiled-route.cjs --hub cli-classifier --prompt "<task>"Follow the returned decision:
route(use itstargets),clarify/defer(disambiguate),reject(refuse). On a{"servingAuthority":"legacy"}sentinel or any error, use the routing below. The front door self-gates on serving-authority. Compiled routing is now the default forcli-classifier. SetSPECKIT_COMPILED_ROUTING=0to force legacy routing fleet-wide. That is the explicit kill-switch.
What ships with it
22 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.
- benchmark/README.md 1.7 KB
- benchmark/reports/README.md 1.9 KB
- canary-cases.v1.json 3.4 KB
- changelog/v1.1.0.0.md 2.4 KB
- description.json 1.4 KB
- graph-metadata.json 5.1 KB
- hub-router.json 2.1 KB
- manifest.cli-classifier.json 195 B
- mode-registry.json 4.2 KB
- playbook-v2/hub-routing/alias-still-resolves.md 6.3 KB
- playbook-v2/hub-routing/judgment-request-routes-to-transport.md 8.0 KB
- playbook/hub-routing/alias-still-resolves.md 5.8 KB
- playbook/hub-routing/deem-request-routes-to-transport.md 4.9 KB
- playbook/hub-routing/jev-request-stays-with-cli-jev.md 4.6 KB
- playbook/hub-routing/judgment-request-routes-to-transport.md 7.5 KB
- playbook/hub-routing/out-of-domain-resolves-nothing.md 4.9 KB
- playbook/manual-testing-playbook.md 7.1 KB
- README.md 4.9 KB
- ROUTER.md 3.5 KB
- serving-closure.manifest.json 4.5 KB
- shared-v3/README.md 408 B
- shared/README.md 742 B
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.
- 16d ago First seen · 163 lines · 41 tokens per session scan A c952315bd70c
cli-classifier is a skill published in the GitHub repository MichelKerkmeester/skilled-agent-harness_spec-driven-loops (37 stars, last pushed yesterday), licensed MIT. It adds 41 tokens to every session and 2,091 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 2026-10-01.
Other skills, from other repositories
agent-platform-model-registry
Agent Platform Model Registry Management. Use when you need to upload, list, describe, update, or delete machine learning models (and their versions) in the Agent Platform Model Registry. Don't use for model training, model deployment to endpoints, or managing non-Agent Platform models.
foundry-config-setup
Resolve missing setup caused by a hardcoded Foundry project endpoint or model in a sample. Use when a sample fails because it uses a placeholder/hardcoded projectendpoint (for example "https://your-project.services.ai.azure.com") or a hardcoded model instead of reading them from the environment.
google-cloud-solution-agentic-analytics-spark-knowledge-catalog
Discovers requirements and designs an end-to-end governed agentic analytics solution using Knowledge Catalog and Managed Service for Apache Spark (Lightning Engine). Use when designing data science and analytics workflows across structured and unstructured distributed data (including in S3, Azure Blob, AlloyDB, and…
training-check
Interactively monitor training metrics from the current Codex session, periodically checking WandB or fallback logs for NaN, divergence, plateaus, and broken runs.
nemo-automodel-launcher-config
Configure NeMo AutoModel job launches for interactive runs, Slurm clusters, and SkyPilot cloud execution.
finetuning
Fine-tune models on Microsoft Foundry using SFT (supervised), DPO (preference), or RFT (reinforcement with graders). Covers dataset preparation, training job submission, deployment, and evaluation. USE FOR: fine-tune, SFT, DPO, RFT, training data, grader, distillation, fine-tuned model, training job, large file…