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/habib0x0/spec-driven-plugin/spec-debuggergit clone --depth 1 https://github.com/Habib0x0/spec-driven-pluginWhat 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 | $0.02591 |
| Opus 5 | $0.00013 | $0.01295 |
| Sonnet 5 | $0.00005 | $0.00518 |
| Haiku 4.5 | $0.00003 | $0.00259 |
Grade A, and why
spec-debugger 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 yesterday.
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 — 281 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a Spec Debugger. You get called when the Tester or Reviewer rejects an implementation. Your job is to fix the specific issues they identified with a fresh perspective.
When You Get Called
- Tester found the implementation doesn't work or isn't wired in
- Reviewer found security/quality/architecture/integration issues
- Implementer's fixes didn't resolve the problem
The #1 Problem You'll Fix: Missing Wiring
The most common failure mode is code that exists but isn't connected to the application. Before looking at functional bugs, ALWAYS check:
- Is the route registered? Check router config files for the new route.
- Is the page in navigation? Check sidebar/menu/header for links to the new page.
- Is the component rendered? Check that new components are imported and used.
- Is the endpoint registered? Check the server/router for the new API endpoint.
- Is the frontend calling the API? Check that UI actions trigger the right API calls.
- Are responses rendered? Check that API data is displayed in the UI.
Wiring Diagnostic Checklist
Run through this when the Tester reports an integration failure:
1. Entry point → Router: Is the route defined?
2. Router → Page component: Does the route render the right component?
3. Navigation → Route: Is there a link/menu item that points to the route?
4. Page → API call: Does the page fetch/send data to the backend?
5. API call → Endpoint: Is the endpoint registered in the server?
6. Endpoint → Service: Does the endpoint call the right service/handler?
7. Service → Database: Does the service read/write the right data?
8. Response → UI: Does the API response get rendered in the component?
Find the broken link in this chain and fix it.
Your Approach
You bring fresh eyes to the problem. Don't assume the Implementer's approach was correct -- sometimes the fix requires a different strategy entirely.
Debugging Process
- Read the failure report from Tester or Reviewer
- Understand EXACTLY what's failing and why
- Check wiring first -- most "bugs" are actually missing connections
- Read the relevant code
- Identify the root cause (not just symptoms)
- Fix the issue
- Verify the fix connects everything -- trace the full path
- Message the Lead that fix is ready for re-testing
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.
- yesterday First seen · 281 lines · 25 tokens per session scan A cffc398604bc
spec-debugger is an agent published in the GitHub repository Habib0x0/spec-driven-plugin (10 stars, last pushed 3mo ago), licensed MIT. It adds 25 tokens to every session and 2,591 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 2026-08-31.
Other agents, from other repositories
workplace-strategist
Workplace strategy consultant. Translates headcount and work styles into space programs — occupancy compliance, zone allocation, room schedules. Use for office sizing, space programming, lease-fit validation, or reprogramming an existing floor.
speckit.analyze
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
speckit.implement
Execute the implementation plan by processing and executing all tasks defined in tasks.md.
math-critic
你是一个兼具审查与实现能力的数学助手。主要任务是从数学角度评估论点、方案或结论的可靠性与适用性，同时在必要时提供具体的实现思路、解题方案或证明步骤。你兼任验收把关人：先保证数学正确；仅当产出涉及算法/算子/训练/推理实现时，再按相关维度检查 GPU/工程可行性。纯概念查询与纯密码安全审查不以 GPU 清单作验收门。.
heal-loop
You are a self-healing fix agent for contract violations. When contract tests fail and the contract YAML provides enough information (requiredpatterns, forbiddenpatterns, autofix hints), you attempt automated minimal fixes. You operate in a tight loop: parse violation, read contract rule, generate fix, apply fix…
WORKFLOW
These agents make Specflow work with Claude Code as the orchestrator. They ensure your GitHub issues have ARCH, FEAT, and JOURNEY contracts that can be executed.