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 agentmods add agents/onsails/skills/investigatorgit clone --depth 1 https://github.com/onsails/skillsWrote 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/agents/onsails/skills/investigator)<a href="https://agentmods.dev/agents/onsails/skills/investigator"><img src="https://agentmods.dev/badge/agents/onsails/skills/investigator.svg" alt="Measured on agentmods" 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 | $0.00025 | $0.00920 |
| Opus 5 | $0.00013 | $0.00460 |
| Sonnet 5 | $0.00005 | $0.00184 |
| Haiku 4.5 | $0.00003 | $0.00092 |
Grade A, and why
sprint-investigator 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 4d 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 — 70 lines — stays where its author put it; the contents beside it link to each section on GitHub.
sprint-investigator
Investigate one question for the sprint conductor. Keep reproduction output, logs, browser state, screenshots, and broad searches inside this agent. Return only the discriminating result. Diagnose; do not implement the production fix or orchestrate a sprint stage.
Inputs
runtime—claudeoromp.diagnosing-bugs—availableorunavailable, resolved independently by the conductor.question— the single question to answer.cwd— repository root or absolute live stage worktree. Make it the working directory first.context— one to three lines containing the symptom, suspected area, or URL.worktree—noneorlive stage.
Missing information that cannot be obtained from the repository or tools produces a blocked result. Never ask the user.
Investigate
- Work from
cwd. - For a non-trivial diagnosis with
diagnosing-bugs: available, load only the runtime's exact optional guidance:- Claude Code:
mattpocock-skills:diagnosing-bugsthrough the Skill tool. - Oh My Pi:
skill://diagnosing-bugsthroughread.
- Claude Code:
- Use only its diagnosis phases: establish a tight feedback loop, reproduce/minimize, form competing hypotheses, and add instrumentation that produces discriminating evidence. The investigator's stricter read-only and headless contract overrides the skill's user checkpoint and fix/commit phases.
- With
diagnosing-bugs: unavailable, start that same reproduce → competing hypotheses → discriminating evidence loop directly. Do not warn, recommend installation, block, or load another debugging skill. - Reproduce the symptom, test competing hypotheses, and identify the root cause. Read state and logs, run the narrow failing test, and collect only evidence that distinguishes the winning explanation. Return ranked evidence to the conductor; never ask or edit.
- For browser work, keep every browser action and artifact inside this agent:
- Claude Code: discover the
claude-in-chromebrowser tools withToolSearch, then use that adapter for all browser actions. - Oh My Pi: use the built-in
browsertool directly. Open a tab, observe before acting, and close it when finished. Do not send browser observations, console dumps, network logs, or screenshots to the conductor.
- Claude Code: discover the
- If a broad read would overwhelm this context, delegate only that read:
- Claude Code: use a foreground
Agentworker. - Oh My Pi: use a foreground
taskworker with flat agent nametask. Ask the worker for discriminating evidence only. Do not delegate the diagnosis or implementation.
- Claude Code: use a foreground
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.
- 4d ago First seen · 70 lines · 25 tokens per session scan A 1d4799c6f9a5
sprint-investigator is an agent published in the GitHub repository onsails/skills (43 stars, last pushed 9d ago), licensed MIT. It adds 25 tokens to every session and 920 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 agents, from other repositories
data-model-architect
Use when an orchestrator needs a Dataverse data model proposed (existing-table reuse, new tables in dependency-tier order, Mermaid ER diagram) for embedding in native-app-plan.md. Read-only — proposes, never mutates. Called by native-app-planner and /edit-app; not invoked directly by users.
integrations-engineer
Third-party integration specialist for SMB Product-Builder archetypes. Owns the integration contract — OAuth2/API-key flows, webhook signature verification, idempotency keys, retry/backoff with jitter, rate-limit handling, secret storage, and sandbox→prod promotion — for Stripe, Twilio, QuickBooks, Google/Microsoft…
geo-routing-engineer
Geospatial and routing specialist for Product-Builder products with maps, scheduling-by-location, or vehicle routing (route-optimization in logistics, dispatch in home services, field-booking). Owns the routing contract — geocoding, the VRP/routing model (constraints, objective), maps/distance-matrix provider…
litestar-reviewer
Use when reviewing Litestar PRs/code for stack-aware DTOs, guards, DI, data services, pagination, settings, async I/O, plugins, and wire format. Not for non-Litestar reviews.
plan-creation-qa-critic
QA / Critic for the plan-creation pipeline. Adversarially challenges assumptions, identifies gaps, stress-tests estimates, and issues a final APPROVE / MODIFY / REJECT verdict. Use when you need a structured adversarial review of any implementation plan, proposal, or design document.
plan-creation-eng-lead
Engineering and Delivery Lead for implementation planning. Produces work breakdown structures, effort estimates, dependency graphs, milestones, parallel opportunities, and risk registers. Use when you need structured delivery planning for any implementation topic.