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 jhlee0409/omni-harness-kit --skill localizegit clone --depth 1 https://github.com/jhlee0409/omni-harness-kitWrote 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/jhlee0409/omni-harness-kit/localize)<a href="https://agentmods.dev/skills/jhlee0409/omni-harness-kit/localize"><img src="https://agentmods.dev/badge/skills/jhlee0409/omni-harness-kit/localize/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/jhlee0409/omni-harness-kit/localize"><img src="https://agentmods.dev/badge/skills/jhlee0409/omni-harness-kit/localize.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 | $0.00118 | $0.00886 |
| Opus 5 | $0.00059 | $0.00443 |
| Sonnet 5 | $0.00024 | $0.00177 |
| Haiku 4.5 | $0.00012 | $0.00089 |
Grade A, and why
localize 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 9d 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.
localize — localize → edit → validate
Do NOT start editing from the first plausible file. Empirically, a staged localize → repair → validate loop resolves real repository issues at a fraction of the cost of broad autonomous exploration (Agentless, arXiv 2407.01489), and the benchmark contract for "resolved" is a change whose tests pass (SWE-bench, arXiv 2310.06770). This skill enforces that discipline: name the target with evidence, make the minimal edit, then prove it with tests.
1. LOCALIZE — produce the target before editing
From the request, find the exact edit site(s). Do NOT edit in this phase.
- Extract the concrete signals from the request: error text, stack trace, symptom, endpoint / field / component name, the file it mentions.
- Search from those signals:
grepthe error string / symbol;globthe module; read the suspected files. For a symbol you will change, run/harness-kit:blast-radius <symbol>to enumerate everything that depends on it, so the edit scope is known up front (not discovered after a half-fix). - If tests exist, use the failing test (or write one that reproduces the symptom) to point at the fault — a failing test IS localization evidence.
Localization artifact (write it down, BLUF):
- Targets: each
file:line+ symbol you intend to change, with the evidence that put it there (the grep hit / the failing test / the reference). - Impact: the blast-radius set for any changed signature (callsites that must move together).
- Uncertainty: what you are NOT sure about + the candidate you'd check next. Non-empty is honest; a confident-but-wrong target is the failure mode here.
2. EDIT — minimal, scoped to the target
Make the smallest change that addresses the request at the localized site. Update every callsite the blast-radius set named — do not leave a half-migrated signature. Match the surrounding idioms.
3. VALIDATE — tests gate the change
- Focused first: run the specific test covering the change (or the reproducer from step 1). It must go from failing → passing.
- Regression next: run the stack's test command (the repo's detected
test_cmd). Report pass/fail counts. - Real-run evidence for a "feature works" claim: a real input → real output,
not just green unit tests (per the kit's
0.3 No premature done). - Record the commands run + outcomes in the handoff / verify artifact.
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.
- 9d ago First seen · 73 lines · 118 tokens per session scan A 25ac92c17dc3
localize is a skill published in the GitHub repository jhlee0409/omni-harness-kit (2 stars, last pushed 1mo ago), licensed MIT. It adds 118 tokens to every session and 886 once invoked, about $0.0006 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 2026-08-31.
Other skills, from other repositories
resolve-plugins
Curate Claude Code skill / MCP / hook picks against live upstream sources and pin them in .claude/settings.json. Adding a process harness (a paired skill set) is an ordinary pick this door handles. Reads stack from docs/techstack.md, workflow signal from docs/overview.md + existing pins, then live-queries six source…
journey-simulation
Use when caller wants to observe how a stranger encounters a flow, artifact, or sandbox — triggers like "simulate a user journey", "test our onboarding / checkout / signup", "will my ICP convert", "how does a cold reader experience this README", "first-time user test", "cognitive walkthrough", or any request to…
check-docs-consistency
Cross-reference project docs for drift, stale references, and contradictions. Outputs timestamped report. Discovers Markdown wherever it lives in the repo.
super-bootstrap
Public entry for the super-bootstrap pipeline — thin orchestrator. Git-inits if absent, then dispatches /super-bootstrap:harness-bootstrap to install or sync the generic runway (always; the runway self-detects fresh-vs-sync). Checks whether seed docs are substantive: greenfield seeds three GAP cards (overview…
release
Prepare a version release — bump version files, commit, and tag. Just run /release with no arguments.
merge
Absorb one or more feature branches into the base branch. Recommends merge vs rebase per branch. On conflict, aborts that branch + surfaces the file list + stops. Resolution out of scope; routes to the harness-named conflict-resolution agent or the user.