AgentPM
AgentPM offers rogrep, a local, open-source Rust binary that indexes coding-agent session history into a searchable, queryable database. It captures conversations, commands, code changes, and outcomes, enabling full-text search, Git/PR trajectory tracing, and structured failure inspection—all without a daemon, account, or network request. The tool includes a bundled skill that teaches AI agents how to search and cite evidence from local history.
- Developer Tools
- Software Only
- Artificial Intelligence
- AI Agents
Funding
Founders
Founder details are not available yet.
Product
Problem
Teams using AI coding agents lack visibility into what those agents actually did, why they made specific decisions, and what state they left the work in. Useful information from past agent sessions—conversations, commands, tests, and outcomes—lives in forgotten chat scrollback, forcing humans to piece together context and future agents to repeat failed approaches or reopen settled questions.
Solution
AgentPM provides rogrep, a local, open-source Rust binary that indexes coding-agent session stores already on a user's machine into a searchable, queryable database. It captures conversations, commands, code changes, decisions, tests, and outcomes, and distills sessions into one-line summaries and typed keyframes linked to the exact turn that produced them. Users can search full text, trace Git branches and PRs, inspect failures by structured facets, and measure usage via repeatable SQLite reports. The tool runs without a daemon, account, or network request, and includes a bundled skill that teaches agents when to search, how to drill down, and how to cite evidence from the local history.
Target Audience
Primary users are developers who need to recover context and keep work moving across agent sessions, engineering leaders who want to see patterns, outcomes, and bottlenecks, and security/governance teams that need auditability and control over agentic software development.
Features
- Automatic indexing of local provider session stores (Claude Code, Codex, Cursor, Grok, Hermes, OpenCode) into a rebuildable Tantivy and SQLite local index
- Full-text search with structured facets, including `tool_status:failed` and `tool_cmd:cargo` for precise failure inspection
- `rogrep trajectory --pr 48` command to rank conversations from captured Git and GitHub operations, tracing branches and PRs
- One-line session summaries that surface repeated friction, useful advice, and improvement opportunities
- Bundled agent skill (`rogrep skill install`) that teaches agents when to search, how to drill down, and how to cite evidence
- Private by construction: no daemon, account, telemetry, analytics, or runtime network request; data written only under its own local data directory
- Installer script, release archives, and manual Cargo install options for Linux and macOS