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 OutlineDriven/odin-claude-plugin --skill reproduce-bug-reportgit clone --depth 1 https://github.com/OutlineDriven/odin-claude-pluginWrote 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/outlinedriven/odin-claude-plugin/reproduce-bug-report)<a href="https://agentmods.dev/skills/outlinedriven/odin-claude-plugin/reproduce-bug-report"><img src="https://agentmods.dev/badge/skills/outlinedriven/odin-claude-plugin/reproduce-bug-report/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/outlinedriven/odin-claude-plugin/reproduce-bug-report"><img src="https://agentmods.dev/badge/skills/outlinedriven/odin-claude-plugin/reproduce-bug-report.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 | $0.00061 | $0.00936 |
| Opus 5 | $0.00030 | $0.00468 |
| Sonnet 5 | $0.00012 | $0.00187 |
| Haiku 4.5 | $0.00006 | $0.00094 |
Grade A, and why
reproduce-bug-report 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 7d 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 — 58 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Reproduce bug report
Contract
| Field | Bound contract |
|---|---|
| Trigger | A UI-visible bug is worth reproducing (a bug report or observed defect exists). |
| Authority | Write only to a named local artifact directory. Roll back any state change that persists after the directory is closed. Do not mutate VCS-tracked files, remote resources, credentials, or data at rest. |
| Side effect | Spawns computer-use repro agents and writes an artifact directory under the named local directory. |
| Done | Structured summary returned with status, steps, environment, evidence, and next step. |
Refusals
- Fixing the bug: use
reproduce-and-fix-issues. This skill reproduces and captures evidence only. - Inferring or hallucinating a bug: rejected. If the report cannot be parsed, return
blocked: no valid bug description. - Mutating VCS-tracked files, remote resources, credentials, or data at rest: rejected.
Inputs
Required:
- Bug report text, error message, observed defect description, or pointer to the issue providing this information.
Optional:
- Environment context (OS, version, terminal, config): supply if available; omit if unknown.
- Reproduction constraints (time budget, retry limit, artifact directory path): use sensible defaults if not supplied.
Procedure
- Parse the bug report: extract the defect description, error message, UI symptoms, and any environment context. Done when: the defect description, symptoms, and environment context are extracted or
blocked: no valid bug descriptionis returned. - Determine the reproduction target: a local environment matching the reported conditions, or a best-effort approximation. Done when: the reproduction target is determined.
- Plan a minimal reproduction: one concrete action sequence that triggers the reported symptom. If no single trigger exists, plan the shortest failing sequence. Done when: a minimal reproduction plan is written.
- Open or confirm an artifact directory for this reproduction session. Name it to identify the bug and the session (e.g.,
repro/<issue-id>/<timestamp>). Done when: the artifact directory is opened. - Execute the minimal reproduction using a computer-use repro agent. Log every action taken and every observable result. Done when: the reproduction is executed with full logging.
- Capture evidence: terminal output, screenshots, logs, error traces, or any artifact produced by the repro attempt. Done when: evidence is captured.
- Compare the observed result against the expected result stated in the bug report. Done when: the comparison is made.
- Record the outcome: reproduced (exact match), partially reproduced (symptom class matches), or not reproduced (no matching symptom). Done when: the outcome is classified.
- Write the structured summary artifact into the artifact directory with Status, Steps, Environment, Evidence, and Next step. Done when: the structured summary is written.
- Close the artifact directory. Roll back any state that persists beyond the directory (e.g., temp files, environment mutations). Done when: the directory is closed and non-directory state is rolled back.
What ships with it
1 file 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.
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.
- 7d ago First seen · 58 lines · 61 tokens per session scan A 3c187a26d41e
reproduce-bug-report is a skill published in the GitHub repository OutlineDriven/odin-claude-plugin (35 stars, last pushed 2d ago), licensed Apache-2.0. It adds 61 tokens to every session and 936 once invoked, about $0.0003 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 2026-09-04.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
insight-error-page
Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…
next-cache-components-optimizer
Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…
next-partial-prefetching-adoption
Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…