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 thangchung/agent-engineering-experiment --skill harness-engineeringgit clone --depth 1 https://github.com/thangchung/agent-engineering-experimentWrote 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/thangchung/agent-engineering-experiment/harness-engineering)<a href="https://agentmods.dev/skills/thangchung/agent-engineering-experiment/harness-engineering"><img src="https://agentmods.dev/badge/skills/thangchung/agent-engineering-experiment/harness-engineering/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/thangchung/agent-engineering-experiment/harness-engineering"><img src="https://agentmods.dev/badge/skills/thangchung/agent-engineering-experiment/harness-engineering.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 | $0.00052 | $0.01602 |
| Opus 5.5 | $0.00021 | $0.00641 |
| Sonnet 5.5 | $0.00010 | $0.00320 |
| Haiku 4.5 | $0.00005 | $0.00160 |
Grade A, and why
harness-engineering 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 — 227 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Harness Engineering
Harness engineering turns repeated coding-agent mistakes into durable repository artifacts:
Harness = Instructions + Constraints + Feedback + Memory + Evaluation + Governance
Use this skill when the user asks to:
- make a repository more reliable for GitHub Copilot or other coding agents
- add durable agent instructions, repository rules, or guardrails
- prevent repeated AI coding-agent mistakes
- record known failure paths and the checks that prevent recurrence
- add lightweight drift checks for project rules
- review, refresh, or update an existing agent harness
Do not use this skill for ordinary feature implementation unless the user asks to improve the repository's agent operating environment.
Core Principles
- Treat the target repository as the source of truth.
- Inspect before editing. Preserve the existing stack, package manager, CI, docs, naming, and architecture.
- Add the smallest useful harness. Prefer updating existing files over adding duplicate guidance.
- Make important rules enforceable where practical through tests, linters, type checks, CI, pre-commit hooks, or drift scripts.
- Use manual review points only when automation would be brittle or misleading.
- Record high-risk failures that should not recur, and name the check or review point that catches recurrence.
- Do not copy generic templates blindly. Adapt every artifact to real evidence in the target repository.
Discovery
Before proposing or making harness changes, inspect the repository for existing rules and evidence.
Read these files and folders when they exist:
README.mdAGENTS.md.github/copilot-instructions.md.github/instructions/.github/workflows/CONTRIBUTING.md- package manifests such as
package.json,pyproject.toml,go.mod,Cargo.toml,pom.xml, orbuild.gradle - existing docs under
docs/ - existing scripts under
scripts/ - existing tests and CI checks
Then summarize:
- stack, package manager, and entry points
- existing development and verification commands
- current agent instructions or repository conventions
- known failures, incidents, flaky paths, or repeated review comments
- gaps where project rules are not enforced
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 · 227 lines · 52 tokens per session scan A 079f32a25c98
harness-engineering is a skill published in the GitHub repository thangchung/agent-engineering-experiment (27 stars, last pushed 5d ago), licensed MIT. It adds 52 tokens to every session and 1,602 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 2026-08-30.
Other skills, from other repositories
evaluators
Author or refine a Phoenix evaluator — code or LLM-as-a-judge — that scores a run's output. Trigger when the user wants to create a new evaluator, improve an existing one's logic or rubric, choose labels, or decide what to measure on a dataset or experiment. Do NOT trigger on: (1) manual prompt drafting (use…
phoenix-client-development
Development guide for the @arizeai/phoenix-client TypeScript SDK — run and resume experiments, manage OpenTelemetry tracer providers with stack-based attach/detach, and write vitest unit and integration tests. Use when adding features to phoenix-client, debugging experiment lifecycle or provider cleanup, modifying…
research-engineer
An uncompromising Academic Research Engineer. Operates with absolute scientific rigor, objective criticism, and zero flair. Focuses on theoretical correctness, formal verification, and optimal implementation across any required technology.
eval-dataset-design
A guide to creating reliable test sets for AI agents. Each evaluation task includes the request, the simulated user's behaviour, a reset starting state, and a success check that can be independently verified.
tika-eval-compare
Compare extracts from two Tika builds over a corpus to detect regressions in content, encoding, exceptions, and embedded-document handling. Use for "compare before/after extracts", "eval this change against the corpus".
cua-driver
Use Cua Driver for desktop or browser tasks that are awkward or unavailable through Bash/APIs, or when the user explicitly wants GUI interaction: app testing, visual bug reproduction, form filling, calendar entry, screenshots, and demo recording. Also covers Cua setup; not OpenAI Codex Computer Use or web research.