Compare newly installed browser skill against existing web/browser tools
Session from pi-mono-sessions on Hugging Face by thomasmustier · MIT · condensed by Coders Talk
Click a segment to jump to that moment
1msession
0interventions
by library · 20h
GOAL
The developer wanted to understand how a newly installed capability compared to existing browser- and web-related skills and extensions already present in the pi-mono repo and various agent skill directories.
- +0Prompt · +0I just installed this. Can we check how it compares to other browser-related and web-related capabilities we have in skills, extensions?
- +0mAgent did · +0mListed repo contents, read README, and grepped packages for 'browser', 'web', and 'extension' references to map out existing capabilities.
- +0mAgent did · +0mRead docs/skills.md and docs/extensions.md to understand pi's skill and extension architecture.
- +0mAgent did · +0mChecked skill directories under .pi/agent/skills, .claude/skills, and .codex/skills to inventory installed skills across agents, hitting a missing-directory error for .pi/agent/skills but continuing with the others.
- +0mAgent did · +0mRead SKILL.md files for browser-tools, brave-search, and web-search to compare their capabilities, descriptions, and use cases.
- +1mAgent did · +1mGrepped example extensions for browser/web references, finding only an unrelated ssh.ts mime-type check.
- +1mOutcome · +1mDelivered a comparison table covering browser-tools (interactive CDP automation), web-search (search+extract with caching), and brave-search (simple search wrapper), noting no relevant web/browser extensions existed.
VERDICT
Next time I'd just name the exact skill I installed up front instead of making the agent infer it from context, since it had to ask which item I meant at the end. The exploratory approach worked fine though - grepping the repo and reading SKILL.md files across the different agent skill directories gave a solid comparison without me needing to intervene.