Distilly
Turns a person's messages, docs, and interviews into a reusable profile that AI agents can load.
- Stars
- 26.3k
- Forks
- 2.2k
- Licence
- MIT
- Language
- Python
- Last update
- 24 days ago
- Type
- Dev tool
What it is
Distilly takes material you provide, such as messages, documents, interviews, or public sources, and builds a Person Profile describing a person's experience, decision patterns, expression, and ways of working. The profile is packaged as an Agent Skill that supported AI agent hosts can install and invoke. The README says it models observable patterns and does not claim to clone the person.
- Build a Work Skill from a departed coworker's chat history and documents
- Create a Persona Skill that captures a mentor's feedback style
- Package a generated profile and install it into a supported agent host
- Use the celebrity research toolchain to organize a public figure's talks and writing
Great for
- Developers who want an agent to answer in the style of a specific colleague or mentor
- People building reusable Agent Skills for supported hosts like Claude Code or Codex
- Users who want a profile of a public figure or author built from public material
- Teams that want to preserve a departing colleague's working knowledge
Not ideal for
- People who need a verified clone of a real person, since the README says it does not claim to clone anyone
- Users without source material, because output quality depends on the materials provided
- Anyone needing a one-click consumer app; the README describes a Skill-based workflow for agent hosts
How to get it
- 1
Clone and build from source
You'll need: Python 3.9+
Clone the repository
git clone https://github.com/titanwings/distilly - 2
Run it
- Clone the repository and review the README for your chosen agent host
- Install the distilly creator Skill into a directory named distilly, as the README specifies
- Provide source material such as messages or documents
- Run the creator workflow to generate a Person Profile
- Install the generated profile into your supported agent host and invoke it
- 3
Get the most out of it
- Use the colleague, relationship, or celebrity family, each with its own collection strategy and dimensions
- Add Lark, DingTalk, or Slack collection for colleague profiles, per the README
- Use the six-dimension research toolchain for public figures, starting from subtitles
- Check the community gallery for existing skills and meta-skills to compare against
Downloads
Strengths
- Supports several agent hosts, including Claude Code, Codex, Hermes, and OpenClaw, per the README
- Output is packaged in the Agent Skill format, so it is portable across supported hosts
- MIT licensed with a large public community and a published technical report
Trade-offs
- The README states the profile reflects observable patterns, not the full person, so results can be incomplete or misleading
- Some hosts such as Grok Bot are only a preview and direct repo installation is not verified
- Quality depends heavily on the amount and quality of input material
How it compares
| Project | Stars | Type | Level | Updated | Best for |
|---|---|---|---|---|---|
| distilly (this) | 26.3k | Dev tool | intermediate | 24 days ago | Developers who want an agent to answer in the style of a specific colleague or mentor |
| ego-lite | 21.3k | App | beginner | 4 days ago | Developers who use Claude Code, Codex, or similar agent CLIs and want them to act in a logged-in browser |
| hermes-agent | 260k | AI model / agent | intermediate | 4 days ago | Developers who want a personal agent they can host on their own server or machine |
| paperclip | 110k | App | advanced | 4 days ago | Developers already running several AI coding agents who want one place to track them |
| Understand-Anything | 89.7k | Dev tool | beginner | 11 days ago | Developers joining a new team who need to learn an unfamiliar codebase |
| agents | 42.1k | Dev tool | beginner | 4 days ago | Developers who already use one of the listed AI coding tools and want ready-made specialist agents |
Stats synced from GitHub. The assessment above was drafted with AI from the project's README on 2026-10-29 and may miss recent changes — check the README before relying on a command. MIT licence: free to use, modify and distribute, including commercially, as long as you keep the copyright notice.