Professor Synapse skill
Professor Synapse is a router skill that summons a matching expert agent for each task and keeps memory across sessions.
- 3,475 ★ on the repository
- Last commit
Extended access
Why this level
- The skill writes memory files and new agents to disk and updates itself from GitHub releases.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/professor-synapseDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The skill writes memory files and new agents to disk and updates itself from GitHub releases.
This is third-party code. Review the repository files before installing.
What it does
Professor Synapse acts as a conductor for a set of expert agents. It clarifies the goal, picks an agent with `summon.py`, then answers in that agent's role. If no expert fits and the need will recur, the skill creates a new one from a template and saves it to the library. A domain researcher that browses the web first and a memory agent are built in. Memory is shared and agent-tagged: SQLite with full-text search, ranked by relevance and recency. A debate mode convenes several specialists for hard decisions. The repository has three flavors: a universal `Prompt.md`, a Claude Desktop skill as a zip and a portable skill for Claude Code and Codex.
Who it is for. Claude and Codex users who want a single entry point for varied tasks and a growing library of their own expert roles with memory.
Good fit when
- The task lies in an unfamiliar field and a specialist role with clarifying questions helps.
- Recurring work should be assigned to a saved expert.
- A decision is contested and several specialists should weigh in.
- Preferences and earlier conclusions must carry over between sessions.
Not a fit when
- The task is narrow and a single command request solves it.
- Memory must not be stored on disk between sessions.
- Self-updating skills that pull from GitHub violate your policy.
Example request
I need a unit economics expert: ask about my business and walk through my model.Limitations
The repository lists no license although the plugin manifest says MIT. The skill updates itself from GitHub releases and writes memory files to disk, so check what lands in `memory/`. `read-gate` hooks block writes to agent folders until the protocol is read. The universal `Prompt.md` is simpler but has no memory or saved agents. The last commit is dated 25 August 2026. Memory files accumulate on disk and are worth reviewing and cleaning by hand. An agent role holds only while the summon hook supports it.
How to disable. Remove the skill in Claude settings or delete its folder from the skills directory. Memory files and created agents live in the skill's data, so delete them separately.
FAQ
How do I install the skill in Claude Desktop?
Download `professor-synapse-claude.zip`, open Settings, then Capabilities and Skills, click Add new skill and upload the archive.
How do I add it to Codex or Claude Code?
Copy the `professor-synapse-skill/` folder into the skills directory, for example `.claude/skills/` or `.codex/skills/`. If the tool keys skills by folder name, rename it `professor-synapse`.
What happens to created agents on update?
An update preserves custom agents, learned patterns and the `memory/` folder. The skill compares its version with the latest release itself.
Can I use it without the skill?
Yes. Paste `Prompt.md` into the system prompt of any chat, but memory and the agent library will not work.
What does the skill memory store?
The user profile and working items live in `memory.json`, long-term memory in the SQLite `longterm.db`. Search uses FTS5 ranked by relevance and recency.
Discussion
No questions or reviews yet.
Related
SkillsThe Maximem Synap skill teaches Claude Code and Codex to add Synap long-term memory to your agent and wire it to the framework
A collection of 14 portable skills for long agent runs: work planning, live app QA, search across past sessions and a phone call
One local memory for Codex, Claude Code, Cursor, Hermes Agent and OpenClaw: agents hand off tasks, review each other and refine their skills overnight
31 skills for Claude Code and Codex: cross-vendor review, task handoff on low quota, a morning brief and a nightly retrospective
Memory for Claude Code: hooks record human and agent decisions, and a grade is assigned afterwards once the outcome is known.
A process kit for agents: a skill router, project memory in a DEV folder and result verification before saying done.
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Repository stars | Main difference |
|---|---|---|---|
| Professor Synapse (this card) | no | 3,475 | |
| agency-agents-zh | no | 22,040 | ships a ready set of roles in Chinese; Professor Synapse creates roles on demand and keeps shared memory |
| Awesome Claude Code Subagents | no | 26,506 | collects Claude Code subagents as a list; Professor Synapse picks and invokes a role for each request |
README in short
The repository offers three routes: a prompt for any AI, a self-building Claude skill and a portable skill. Both skills keep agents, patterns and shared memory.