noui generalize

AI Community

Skill notes

Works for both MCP and Skill outputs

The generalization logic is identical for either output format: MCP output: edit workbench/mcp_servers/<app>/<server>/operations/<tool>.py — the body of async def execute(...). The FastMCP tool signature in server.py mirrors the execute() signature, so renaming params also requires updating the server.py decorator args.

Critical Rules (Never Violate)

NEVER rewrite all tools at once — propose the new name and parameter list for each tool and get user approval before editing any files ALWAYS preserve existing execution mechanics (URL, method, headers, CDP vs. httpx) unless deliberately switching modes — only the Python function interface changes ALWAYS hardcode values that were static in the recording (session routing params, build labels, bl, f_sid, reqid, soc_app, etc.) — do not expose infrastructure params to the caller

Phase 0 — Diagnose Execution Strategy

Before touching tool names or params, check whether the tools actually work.

0a. Find and test the server

Run a quick end-to-end test:

0b. Classify the result

Result · Diagnosis · Next step 200 with real data · Tool works — skip to Phase 2 (interface cleanup) No Tabby page matching '<domain>' · No tab open on the target site · Open the site in the Tabby browser, or tabby session ensure --profile <id> Connection refused on 9222 · Tabby session not running · tabby session ensure --profile <id> CORS error / credentials include not allowed · Cross-origin fetch blocked · Re-export with --execution-mode http

Fluxly profile Runtime and permissions are inferred by Fluxly static checks, not author declarations or runtime test results.
Domain
AI
Author-declared agents
No explicit declaration found; this is not inferred or tested compatibility
Static check
88 / 100 · heuristic scan, not runtime safety proof
Runtime evidence
Not run; no isolated runtime test evidence is currently recorded
Author / version / license
@tomevault-io · no license declared
Token usage
Heavy
Setup complexity
Manual integration
External API key
Required · Vendor-specific
Operating systems
Docker
Runtime requirements
Python · Docker
Detected file/system behavior
  • Read-only
  • Write / modify
  • Shell exec
Network behavior
External requests
Install commands
None (reference only)

Heads up: Tool permissions are not declared; actual access depends on the runtime.

Output preview noui-generalize.preview
# Works for both MCP and Skill outputs

- **MCP output:** edit `workbench/mcp_servers/<app>/<server>/operations/<tool>.py` — the body of `async def execute(...)`. The FastMCP tool signature in `server.py` mirrors the execute() signature, so renaming params also requires updating the `server.py` decorator args.
- **Skill output:** edit `workbench/skills/<app>/operations/<tool>.py` — the body of `async def execute(...)` *and* the `_build_parser()` argparse registrations, because the skill operation is a standalone CLI script. The `SKILL.md` body's command examples also reference the flag names, so update them too (or regenerate SKILL.md by re-running `workflow export --as skill --description-override "..."`).

Discussion

Powered by GitHub Discussions. Sign in with GitHub to comment, react, or subscribe.