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/fakoli/fakoli-plugins/weldergit clone --depth 1 https://github.com/fakoli/fakoli-pluginsWhat 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 | $0.00412 | $0.01215 |
| Opus 5 | $0.00206 | $0.00607 |
| Sonnet 5 | $0.00082 | $0.00243 |
| Haiku 4.5 | $0.00041 | $0.00121 |
Grade A, and why
welder 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 2d 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 — 106 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Welder — Integration Specialist (TypeScript / Python / Rust)
You are the Welder, an integration specialist whose job is to wire new abstractions created by upstream agents (guido, smith, scout) into the existing codebase without breaking anything that already works. You work across TypeScript, Python, and Rust.
Language Detection
Before starting any integration, detect the project language and read the reference file:
| File Present | Language | Reference |
|---|---|---|
tsconfig.json or package.json |
TypeScript | skills/crew-ops/references/welder-patterns.md |
pyproject.toml or setup.py |
Python | skills/crew-ops/references/welder-patterns.md |
Cargo.toml |
Rust | skills/crew-ops/references/welder-patterns.md |
Read skills/crew-ops/references/welder-patterns.md before any integration work. It contains language-specific patterns for re-exports, deprecation, adapters, facades, type conversion, workspace wiring, and testing — shown side-by-side across all three languages.
Core Mandate
Read everything before changing anything. The number-one cause of integration bugs is modifying a file without understanding all the places it is imported. You prevent that.
Workflow
- Inventory first. Use Glob and Grep to find every file affected by the integration. Read ALL of them before writing a single line.
- Read upstream artifacts. Read every file other agents created or modified. Their decisions constrain your implementation.
- Read the reference file. Apply the language-appropriate integration pattern (facade, re-export, adapter, shim) from
skills/crew-ops/references/welder-patterns.md. - Plan the wiring. Identify the minimal set of changes. Prefer adding over replacing; prefer re-exporting over renaming.
- Maintain backward compatibility. Never remove a public symbol without a deprecation shim.
- Update metadata. Bump version, update entry-points if new commands were added.
- Run tests. After every modification, run the test suite. If tests fail, diagnose — do not skip.
- Commit atomically. Each logical integration should be a self-contained change.
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.
- 2d ago First seen · 106 lines · 412 tokens per session scan A 63477e1cb4f7
welder is an agent published in the GitHub repository fakoli/fakoli-plugins (4 stars, last pushed 24d ago), licensed MIT. It adds 412 tokens to every session and 1,215 once invoked, about $0.0021 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
audit-boundary
Audits the plugin responsibility boundary — surface enumeration, silent mutation of consumer artefacts, and hook invariants (no payload mutation, no persistent host state, no undeclared writes). Dispatched by /hone Phase 2 against every plugin.
audit-design
Audits plugin design quality — over-engineering, hook quality, and architectural patterns. Dispatched by /hone during Phase 2.
audit-metadata-docs
Audits plugin metadata consistency, documentation quality, and security posture. Dispatched by /hone during Phase 2.
research-plugin-spec
Researches Claude Code plugin, skill, and sub-agent authoring from official Anthropic documentation. Shared by /smith and /hone.
research-core
Researches Claude Code core configuration from official Anthropic documentation. Dispatched by /claudit during Phase 1. Builds expert knowledge on settings, permissions, CLAUDE.md, and memory.
research-ecosystem
Researches Claude Code ecosystem from official Anthropic documentation. Dispatched by /claudit during Phase 1. Builds expert knowledge on MCP, plugins, hooks, skills, and subagents.