pact-researcher

pact-researcher is an agent for Claude Code from jonathanmr22/pact. It costs 81 tokens per session (1,453 once invoked), scanned A, original, MIT.

A research agent that checks whether information about a package, library, API, security pattern, or domain topic is already verified. It reads project knowledge files and saves confirmed findings for later use.

In plain words
What is it for?
Use it before using an unfamiliar package, library, or API, when building a new service, or when investigating unexpected dependency behavior.
Why use it?
It helps prevent code from being written based on guesses about dependencies, external services, or technical patterns.

Agent for Claude Code

Written for Claude Code: shipped in a Claude Code plugin. Also seen: model in frontmatter.

Part of the pact plugin — 5 skills, 3 agents shipped together

Install

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.

agentmods
npx agentmods add agents/jonathanmr22/pact/pact-researcher
Clone the repo
git clone --depth 1 https://github.com/jonathanmr22/pact

Made for: Claude Code.

Or install pact, the plugin that ships this one along with the rest of its 5 skills, 3 agents.

Wrote 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.

agentmods badge for pact-researcher

README.md
[![agentmods](https://agentmods.dev/badge/agents/jonathanmr22/pact/pact-researcher.svg)](https://agentmods.dev/agents/jonathanmr22/pact/pact-researcher)
Your own site
<a href="https://agentmods.dev/agents/jonathanmr22/pact/pact-researcher"><img src="https://agentmods.dev/badge/agents/jonathanmr22/pact/pact-researcher.svg" alt="Measured on agentmods" height="20"></a>
Per session 81 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,453 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00081 $0.01453
Opus 5 $0.00041 $0.00727
Sonnet 5 $0.00016 $0.00291
Haiku 4.5 $0.00008 $0.00145

Measured 6d ago against content hash cf3cf19cb12b, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-05, from the pricing page.

Security

Grade A, and why

pact-researcher 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 6d 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.

plugins/pact/agents/pact-researcher.md · 161 lines

How it starts

The opening of the file, as written. The whole thing — 161 lines — stays where its author put it; the contents beside it link to each section on GitHub.

PACT Researcher — Knowledge Compound Agent

You are PACT's research specialist. Your job is to answer ONE question: "Do we actually know this, or are we guessing?"

You receive a topic — a package name, an API, a design pattern, a domain question. You return verified knowledge and save it for future sessions. The main session's context window stays clean while you do the deep work.

Phase 0: Read Project Context

Before doing anything else, read .claude/pact-context.yaml if it exists. This tells you what project you're in, what stack it uses, what external services it integrates with, and where knowledge files live. Use this to:

  • Skip researching things that are already documented in external_services
  • Know which package manager, framework, and patterns apply
  • Find the knowledge directory path (may not be the default)

If the file doesn't exist, proceed without it — but note that your findings may be less project-specific.

Your Process

Phase 1: Check What We Already Know

  1. Read knowledge/KNOWLEDGE_DIRECTORY.yaml — search for tags matching your topic. This is the cross-system index.

  2. If package-related: read knowledge/packages/{name}.yaml if it exists. Check the verified_date — if recent and covers your question, return it directly. No new research needed.

  3. If domain/architecture-related: scan knowledge/research/ entries with matching tags. A previous session may have already synthesized exactly what you need.

  4. Check _SOLUTIONS.yaml — if the topic relates to a bug pattern, a graduated solution may already exist.

Phase 2: Research (only if Phase 1 didn't answer the question)

  1. Project-level research — read the actual source code. Grep for usage patterns. Read the relevant feature flow docs. Understand how THIS project uses the thing in question.

  2. External research — WebSearch for official documentation, API references, GitHub issues, known gotchas. Prefer primary sources (official docs, source repos) over blog posts and tutorials.

Read the full file on GitHub · 161 lines

Changes

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.

  1. 6d ago First seen · 161 lines · 81 tokens per session scan A cf3cf19cb12b

Subscribe to this mod's changes

pact-researcher is an agent published in the GitHub repository jonathanmr22/pact (15 stars, last pushed 2mo ago), licensed MIT. It adds 81 tokens to every session and 1,453 once invoked, about $0.0004 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 2026-08-30.

Related

Other agents, from other repositories

agent-system

An AgentSystem composes multiple Agents into a directed graph that Orloj executes as a coordinated workflow. The graph defines how messages flow between agents during task execution.

OrlojHQ/orloj · 0 tokens

board-workflow

This document describes the end-to-end workflow for AI-assisted issue resolution, from initial issue creation through merged PR.

AndrewAltimit/template-repo · 0 tokens

claude-expression

DISCLAIMER: This document is managed exclusively by Claude Code. The repository admin (AndrewAltimit) does not manage this file and is not allowed to directly edit it. Any updates must come from Claude Code through collaborative sessions.

AndrewAltimit/template-repo · 0 tokens

codebase-analysis-pipeline

This document describes a two-stage automated pipeline for continuous codebase improvement.

AndrewAltimit/template-repo · 0 tokens

gemini-expression

DISCLAIMER: This document is managed exclusively by Gemini. The repository admin (AndrewAltimit) does not manage this file and is not allowed to directly edit it. Any updates must come from Gemini through code review sessions or collaborative agent interactions.

AndrewAltimit/template-repo · 0 tokens

plan-creation-architect

Technical architect for implementation plan creation. Analyzes system design, component decomposition, integration points, design patterns, and technical trade-offs. Reads Product Owner output and optional research synthesis. Use when architectural analysis is needed for a new feature, system, or implementation plan.

QBall-Inc/the-bulwark · 59 tokens