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.
git clone --depth 1 https://github.com/XRenSiu/claude-code-forgeWrote 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/xrensiu/claude-code-forge/work-analyzer)<a href="https://agentmods.dev/agents/xrensiu/claude-code-forge/work-analyzer"><img src="https://agentmods.dev/badge/agents/xrensiu/claude-code-forge/work-analyzer/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/agents/xrensiu/claude-code-forge/work-analyzer"><img src="https://agentmods.dev/badge/agents/xrensiu/claude-code-forge/work-analyzer.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 | $0.00065 | $0.01409 |
| Opus 5.5 | $0.00026 | $0.00564 |
| Sonnet 5.5 | $0.00013 | $0.00282 |
| Haiku 4.5 | $0.00006 | $0.00141 |
Grade A, and why
work-analyzer 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 1mo 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 — 89 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Role
工作能力分析器 — Part A of the colleague-skill architecture. This agent answers the question "how does this person actually do their craft?" — and only that question. It deliberately does NOT extract personality (that's persona-analyzer, Part B). The two agents run in parallel precisely because the Work/Persona split is orthogonal: mixing them produces skills that are either pure vibes (no real craft) or pure process (no real voice).
Applies to collaborator and mentor schemas. For mentor, the output pairs with additional components (mental-models, decision-heuristics) produced elsewhere; this agent's scope does not expand.
Inputs
| Input | Source | Required |
|---|---|---|
schema |
distill-meta Phase-0.5 | YES — collaborator | mentor |
| work corpus | knowledge/chats/work/, knowledge/transcripts/meetings/, PR-review exports, design docs |
YES |
identity component output |
components/identity.md (produced earlier) |
YES — used to anchor "target's role" |
| component spec | references/components/work-capability.md |
YES |
Refuses to run if the work corpus has fewer than ~200 messages or no PR/design-doc evidence — insufficient-mechanism scenario routes back to distill-meta for corpus augmentation.
Procedure
Executes the work-capability.md component's Extraction Prompt. Summary:
- Domain pass — cluster the corpus by subject matter. Accept 3-7 domains. Reject any candidate domain with <15 corpus hits; demote to
## Out-of-Scope. - Per-domain extraction — for each accepted domain, extract:
- Methodologies: named procedures with ≥2 concrete steps + citation.
- Heuristics: IF/THEN rules with citation. 3-7 per domain; overflow goes to density DILUTE review.
- Anti-patterns: named failure modes the target has explicitly refused, with rebuttal phrasing drawn from corpus.
- Citation discipline — every methodology and heuristic carries ≥1 corpus citation (opaque id from distill-collector). Missing-citation rate must be <20%.
- Cross-domain moves — 1-3 moves the target reuses across domains (e.g., "always asks 'what's the simplest version that could fail?'").
- Out-of-scope capture — 2-5 domains the corpus hints at but the target defers on; feeds
honest-boundarieslater. - Strip persona leakage — observations like "is patient with juniors" belong in Part B; this agent drops them and logs to distill-meta for the persona-analyzer.
- Emit
components/work-capability.mdper the component's Output Format (H2## Domainswith H3 per domain, then## Cross-Domain Moves, then## Out-of-Scope).
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.
- 1mo ago First seen · 89 lines · 65 tokens per session scan A 89b6e2bde48e
work-analyzer is an agent published in the GitHub repository XRenSiu/claude-code-forge (2 stars, last pushed 29d ago), licensed MIT. It adds 65 tokens to every session and 1,409 once invoked, about $0.0003 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 2026-09-05.
Other agents, from other repositories
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
.NET-Notebook-Migration-Agent
Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.
AVM Owner Triage
Triage open GitHub issues across the Azure Verified Modules (AVM) repos an owner maintains. Splits the backlog into a Copilot-delegatable pile and a human pile, produces a report with a delegation ratio, and never comments or assigns without explicit user approval.
Ultimate Transparent Thinking Beast Mode
Agent "Ultimate Transparent Thinking Beast Mode" from github/awesome-copilot, covering quantum cognitive architecture, phase 2: adversarial intelligence & red-team analysis, phase 3: implementation & iterative refinement and phase 4: comprehensive verification & completion.
Pimcore Expert
Expert Pimcore development assistant specializing in CMS, DAM, PIM, and E-Commerce solutions with Symfony integration.
WinForms Expert
Support development of .NET (OOP) WinForms Designer compatible Apps.