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 burin-labs/harn --skill harn-de-slopgit clone --depth 1 https://github.com/burin-labs/harnWrote 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/burin-labs/harn/harn-de-slop)<a href="https://agentmods.dev/skills/burin-labs/harn/harn-de-slop"><img src="https://agentmods.dev/badge/skills/burin-labs/harn/harn-de-slop/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/burin-labs/harn/harn-de-slop"><img src="https://agentmods.dev/badge/skills/burin-labs/harn/harn-de-slop.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 | $0.00022 | $0.01087 |
| Opus 5 | $0.00011 | $0.00544 |
| Sonnet 5 | $0.00004 | $0.00217 |
| Haiku 4.5 | $0.00002 | $0.00109 |
Grade A, and why
harn-de-slop 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 — 130 lines — stays where its author put it; the contents beside it link to each section on GitHub.
De-slop Harn changes
Use this skill to remove accidental complexity without shrinking the durable outcome. The goal is ambitious behavior behind boring seams.
Pair it with [[harn-orchestration]] for execution ownership, [[harn-product-quality]] for cross-surface behavior, and [[harn-testing]] for evidence.
Start with the outcome
- State the user or operator outcome in one sentence.
- Name a plausible falsifier.
- Identify the module that should own the behavior.
- Identify every current copy or projection.
- Distinguish class-killing cleanup from unrelated speculative cleanup.
- Preserve launch requirements while simplifying the implementation.
- Do not use "simpler" to justify a smaller product.
One owner
- A semantic decision gets one authoritative owner.
- Make other surfaces adapters or generated projections.
- Delete parallel scheduling, lifecycle, policy, and completion logic.
- Fix the owner or its interface when projections disagree.
- Prefer the existing Harn runtime and stdlib over host glue.
- Keep hosts responsible for native UX, mutations, and undo.
- Keep Harn Cloud responsible for remote durability, not language semantics.
Deep modules
- Prefer a small interface hiding substantial behavior.
- Apply the deletion test: if removing the module spreads complexity across callers, the module is earning its keep.
- Reject pass-through wrappers that rename parameters without hiding policy.
- Put a seam where behavior actually varies.
- One adapter is a hypothetical seam; two justified adapters make it real.
- Keep internal test seams private to the implementation.
- Test through the same interface callers use.
Closed contracts
- Replace open dictionaries with named records or enums.
- Use
Result<T, E>for fallible operations. - Give public functions explicit return types.
- Validate untrusted input once at the owning seam.
- Preserve diagnostic codes and actionable spans.
- Encode lifecycle and completion as typed states.
- Encode evidence and receipts as data, not prose conventions.
- Avoid
anyas a compatibility blanket.
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 · 130 lines · 22 tokens per session scan A 53f2d3bfce48
harn-de-slop is a skill published in the GitHub repository burin-labs/harn (22 stars, last pushed today), licensed Apache-2.0. It adds 22 tokens to every session and 1,087 once invoked, about $0.0001 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
commit
Auto-fix, lint, test, and commit changes with a conventional commit message.
fix
Auto-fix all code quality issues with rector, cs-fixer, and phpstan.
archestra-pr-hygiene
Use when opening or updating an Archestra pull request to write an accurate conventional title and a practical, evidence-based description.
code-review
A code-review and quality-analysis skill written in Chinese. It is intended to inspect code, find potential problems, and suggest improvements.
translate
A translation tool for converting user-provided text between languages, with Chinese and English as the default pair.
migration-review
Review database migration files when a change adds or modifies paths under migrations/. Use it before merge to collect forward, rollback, locking, and data-safety evidence.