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 humanerd-drew/opencode-drewgent --skill external-tool-evaluationgit clone --depth 1 https://github.com/humanerd-drew/opencode-drewgentWrote 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/humanerd-drew/opencode-drewgent/external-tool-evaluation)<a href="https://agentmods.dev/skills/humanerd-drew/opencode-drewgent/external-tool-evaluation"><img src="https://agentmods.dev/badge/skills/humanerd-drew/opencode-drewgent/external-tool-evaluation.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 5.1 | $0.00062 | $0.02793 |
| Opus 5 | $0.00031 | $0.01396 |
| Sonnet 5 | $0.00012 | $0.00559 |
| Haiku 4.5 | $0.00006 | $0.00279 |
Grade A, and why
external-tool-evaluation scanned grade A with 1 finding 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 3d 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
description: Evaluate a third-party GitHub repo or external tool for {{AGENT_NAME}} integration. Use raw markdown curl + GitHub API to bypass UI, identify algorithm taxonomy, map to {{AGENT_NAME}} hot spots, score 3-4 in How it starts
The opening of the file, as written. The whole thing — 154 lines — stays where its author put it; the contents beside it link to each section on GitHub.
External Tool Evaluation — {{AGENT_NAME}} Integration Workflow
When to use
User shares a GitHub repo link OR asks "what about X for {{AGENT_NAME}}" / "should we use X" / "evaluate X for integration." Trigger keywords: "토큰 똑똑하게", "이 도구 쓸 수 있을까", "X 통합 검토", "headroom / RTK / lean-ctx / kompress / X 평가해줘".
Why a skill is needed
GitHub UI via mcp_browser_navigate returns ~700-element accessibility tree with 95% chrome. Raw markdown via curl gives the actual content in 1 call. Repo root docs/ is often a Next.js site, not source — docs/spec/ or crates/ is where architecture lives. Without this method, the agent spends 5+ browser tool calls extracting what curl | head returns in 1.
Workflow (7 steps)
1. Raw markdown fetch (bypass GitHub UI)
curl -sL https://raw.githubusercontent.com/{owner}/{repo}/main/README.md | head -200
# 4.3k stars + last-commit 3h ago = maturity check satisfied
2. Repo health snapshot
curl -sL https://api.github.com/repos/{owner}/{repo}
# Check: stars, last_push_at (NOT updated_at), default_branch, license
3. Find actual architecture/spec files
# root docs/ is often a Next.js site — list subdirs
curl -sL https://api.github.com/repos/{owner}/{repo}/contents/docs
# For Rust+Python repos: look in crates/, headroom/transforms/, sdk/
# For docs-heavy repos: docs/spec/, docs/content/spec/
Pitfall: docs/README.md is often the website's README (Vercel deploy info), not the source README.
4. Identify the algorithm/feature taxonomy
Read 2-3 spec files or core source files. Extract:
- What it does (1-line)
- Algorithm list (each with 1-line description)
- Integration modes (proxy / library / wrap / MCP / middleware)
- Benchmarks (savings %, accuracy preserved)
- Reversibility (can original be retrieved?)
Special category: Agent behavior rulesets. Some tools (ponytail, caveman, etc.) don't have traditional algorithms — they ship a ruleset/prompt that modifies how the agent thinks before writing code. For these, extract:
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.
- 3d ago First seen · 154 lines · 62 tokens per session scan A e6bb814a1424
external-tool-evaluation is a skill published in the GitHub repository humanerd-drew/opencode-drewgent (2 stars, last pushed 1mo ago), licensed MIT. It adds 62 tokens to every session and 2,793 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
inno-prepare-resources
Loads the evaluation instance, searches GitHub for related repositories, builds a dataset description, queries the Prepare Agent for reference codebases, and downloads arXiv paper sources.
github-official-mcp
Official GitHub MCP Server, by GitHub. Provides seamless integration with GitHub APIs, enabling advanced automation and interaction capabilities for developers and tools.
orch-research
Research top repos for a user-supplied topic with expert panel evaluation and presentation packaging.
pilot-github-bridge
Bridge GitHub webhook events as Pilot Protocol events. Use this skill when: 1. You need to receive GitHub events in Pilot agents 2. You want to trigger agent actions on repository events 3. You're building CI/CD workflows with Pilot agents Do NOT use this skill when: - You need direct GitHub API access (use gh CLI…
github-skill-forge
A tool for turning any GitHub repository into a standard Trae skill for an AI coding agent.
langsmith-observability
LLM observability platform for tracing, evaluation, and monitoring. Use when debugging LLM applications, evaluating model outputs against datasets, monitoring production systems, or building systematic testing pipelines for AI applications.