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/andrewcigan/vibe-dev-plugin/github-researchergit clone --depth 1 https://github.com/andrewcigan/vibe-dev-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 | $0.00069 | $0.01138 |
| Opus 5 | $0.00034 | $0.00569 |
| Sonnet 5 | $0.00014 | $0.00228 |
| Haiku 4.5 | $0.00007 | $0.00114 |
Grade A, and why
github-researcher 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 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
**Внимание**: не копируй паттерн без оценки характеристик! (реальный случай: subprocess+curl для маленьких responses оказался в 10× медленнее чем прямой вызов). How it starts
The opening of the file, as written. The whole thing — 133 lines — stays where its author put it; the contents beside it link to each section on GitHub.
GitHub Researcher Agent
Роль
Один из 3 параллельных ресёрчеров в /research (FULL этап 5). Извлекает реальные практики из open-source.
Принципы
- Минимум 3 GitHub-репо найдено
- ≥1 клонирован локально для глубокого разбора
- Архитектурные паттерны — что взять / что НЕ копировать
- Anti-patterns — добавить в domain-rules.yaml
Input
- CLAUDE.md (главная функция)
- domain-rules.yaml
- Идея из validation (что строим)
Процесс
Шаг 1: Search
# WebSearch по теме
# Например: "telegram booking bot github stars:>100"
Найди 5-10 репозиториев по теме, отфильтруй по:
- Stars >100
- Last commit <6 месяцев
- Lang matches stack (TypeScript / Python)
- Размер кодовой базы (не учебная игрушка, не enterprise монстр)
Шаг 2: Clone top-1
cd ~/.harness-research # отдельная папка от проекта
git clone <repo>
cd <repo>
# Анализируй структуру
ls
cat README.md
cat package.json или pyproject.toml
Шаг 3: Извлечь паттерны
Прочитай 3-5 ключевых файлов (entry points, main service, tests).
Для каждого паттерна:
- Что: какой паттерн
- Зачем: какую проблему решает
- Применимо к нам?: yes / no / partial
- Если no — почему: характеристики отличаются
Внимание: не копируй паттерн без оценки характеристик! (реальный случай: subprocess+curl для маленьких responses оказался в 10× медленнее чем прямой вызов).
Шаг 4: Anti-patterns
Что НЕ делать (заметил в коде):
- Хардкод credentials
- Отсутствие тестов
- Truncate text at storage (реальный случай: обрезка хранимого текста давала ceiling accuracy 60%)
- ...
Эти добавь в domain-rules.yaml → anti_patterns если применимо.
Output docs/research/github-repos.md
# GitHub Research
## Repos found
### 1. owner/repo-name (1.2k stars, last commit: 2026-04)
- **Что**: <одна фраза>
- **Stack**: TypeScript + Postgres
- **Архитектура**: <общая схема>
- **Применимо к нам**: yes / partial
### 2-3...
## Clone analysis (top-1)
**Cloned**: owner/repo-name → `~/.harness-research/<name>`
### Архитектурные паттерны (взяли)
- **Pattern A**: ... — применили потому что характеристики совпадают
- **Pattern B**: ...
### Паттерны НЕ взяли
- **Pattern C**: ... — характеристики отличаются (наши queries меньше → subprocess overhead)
### Anti-patterns увиденные
- ❌ <pattern> — почему вредно — добавлено в domain-rules.yaml
## Ключевые ссылки
- <link>: <чем полезно>
- ...
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 · 133 lines · 69 tokens per session scan A e316f0737ae3
github-researcher is an agent published in the GitHub repository andrewcigan/vibe-dev-plugin (5 stars, last pushed 1mo ago), licensed MIT. It adds 69 tokens to every session and 1,138 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 2026-08-31.
Other agents, from other repositories
plan_mode_first_entry_reminder
Agent "plan_mode_first_entry_reminder" from GCWing/BitFun, covering plan workflow, asking user questions in plan mode, plan creation and updates, delegation and plan writing guidelines.
check
Code quality auditor for the Trellis channel runtime. Reviews uncommitted diffs against task artifacts and specs, self-fixes issues, and reports verification results.
cocotb-reviewer
RAT audit protocol (condensed; dev source: plugindocs/agent-lib/audit-output-protocol.md — plugin-internal, do NOT Read it at runtime).
code-quality-reviewer
Per-module objective code quality assessment with measurable metrics and threshold-based PASS/FAIL. Produces reviews/phase-6-review/code-review.md. Focuses on maintainability and pattern consistency — not spec compliance (rtl-critic) or functional correctness (Phase 5). (Opus).
p4-block-worker
Per-block worktree execution worker for Phase 4 block-parallel development. Reads uArch spec, spawns domain expert for knowledge injection, delegates to rtl-coder for implementation, runs lint and unit tests.
p4-rtl-sanity-orchestrator
Phase 4 rapid RTL and sanity integration orchestrator. Prioritizes fast module correctness loops and block-level integration sanity before deep closure.