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 wahidyankf/ose-public --skill harness-compatibility-protocolgit clone --depth 1 https://github.com/wahidyankf/ose-publicWrote 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/wahidyankf/ose-public/harness-compatibility-protocol)<a href="https://agentmods.dev/skills/wahidyankf/ose-public/harness-compatibility-protocol"><img src="https://agentmods.dev/badge/skills/wahidyankf/ose-public/harness-compatibility-protocol/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/wahidyankf/ose-public/harness-compatibility-protocol"><img src="https://agentmods.dev/badge/skills/wahidyankf/ose-public/harness-compatibility-protocol.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 | $0.00033 | $0.01187 |
| Opus 5.5 | $0.00013 | $0.00475 |
| Sonnet 5.5 | $0.00007 | $0.00237 |
| Haiku 4.5 | $0.00003 | $0.00119 |
Grade A, and why
harness-compatibility-protocol 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 2d 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 — 77 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Repository Harness Compatibility Protocol
Overview
Two agents share one taxonomy: deterministic and semantic Phase 0 parity invariants (offline, Bash-based) and seven Phase 1 external-drift dimensions (web-research-backed). The checker detects; the fixer remediates what's safely mechanical and flags the rest for human judgment.
Reference Modules
- phase0-parity-invariants.md — the five invariants, each with detection tool/pass/fail/criticality AND fix scope (auto-fixable vs. human-required)
- phase0-quality-gate-filter.md — exact-ID delegation and retained semantic parity for quality-gate invocation
- phase1-drift-dimensions-d1-d3.md and phase1-drift-dimensions-d4-d7.md — the seven dimensions (D1–D7), each with drift indicator/criticality AND fix target/action
- checker-workflow.md and checker-finding-format.md — the checker's own workflow steps, research delegation pattern, and finding format
- fixer-confidence-and-scope.md, fixer-patterns-and-process.md, and fixer-report-format.md — the fixer's own confidence re-validation, fix patterns, process summary, fix report format, and FALSE_POSITIVE carry-forward
Core Principles
- Phase 0 leaves declared tools their properties. Under the quality gate, vendor terms and
adapter generation belong to
governance-vendorandharness-adapters, never to findings; unregistered semantic parity still runs. - Only Invariant 3 (binding sync) and most Phase 1 dimensions auto-fix — anything touching governance prose, root-instruction files, agent-set divergence, or a new color/tier mapping requires human judgment.
- Confidence propagates from the checker's cited source:
[Verified]→ HIGH,[Needs Verification]/[Unverified]→ MEDIUM (fixer downgrades and skips),[Outdated]→ FALSE_POSITIVE. - Conservative drift threshold — flag substantive changes only (a different filename, a renamed directory, a removed required field), never minor wording differences.
What ships with it
12 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.
- README.md 391 B
- reference/checker-finding-format.md 1.3 KB
- reference/checker-workflow.md 3.0 KB
- reference/fixer-confidence-and-scope.md 1.3 KB
- reference/fixer-patterns-and-process.md 2.3 KB
- reference/fixer-report-format.md 1.3 KB
- reference/phase0-invariant-6-config-parity.md 2.9 KB
- reference/phase0-parity-invariants.md 3.4 KB
- reference/phase0-quality-gate-filter.md 630 B
- reference/phase1-drift-dimensions-d1-d3.md 1.8 KB
- reference/phase1-drift-dimensions-d4-d7.md 2.1 KB
- reference/README.md 1.4 KB
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.
- 2d ago First seen · 77 lines · 33 tokens per session scan A dda452f4e35f
harness-compatibility-protocol is a skill published in the GitHub repository wahidyankf/ose-public (12 stars, last pushed today), licensed MIT. It adds 33 tokens to every session and 1,187 once invoked, about $0.0001 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 2026-10-14.
Other skills, from other repositories
investigate-cloudflare
Investigate deployed Worker failures, regressions, and missing operational signals using Cloudflare Issues, logs, traces, SQL, or Observability MCP. Produce bounded evidence and verify recurrence against the deployed version.
cli-wiring
Bug class: Commands implemented in packages/squad-cli/src/cli/commands/ but never routed in cli-entry.ts.
investigate
Investigate a single EmDash issue end to end -- classify, choose a repro path, reproduce, diagnose, verify, and (only on an explicit maintainer fix directive) fix and confirm. Every verdict carries its evidence.
repro-admin
Reproduce an EmDash admin UI bug. Attach a container, start the demo dev server, drive the admin with agent-browser using the dev-bypass session, and capture the reproduction as screenshots plus a replayable transcript.
repro-public
Reproduce a bug in the public-facing rendered site (not the admin). Attach a container, start the demo dev server, drive public routes with agent-browser, and capture the reproduction as screenshots plus a replayable transcript.
verify
Decide whether the diagnosed behaviour is actually a bug or the code doing what it was designed to do. This is the gate that guards the fix stage.