pend_skill
- Python
969
GitHub Stars
2
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill bfly123/claude_code_bridge --skill pend- SKILL.md640 B
- SKILL.md.powershell646 B
Overview
This skill lets you quickly view the latest reply from a chosen AI provider (Gemini, Codex, OpenCode, Droid, or Claude) from the command line. It focuses on fast retrieval with minimal overhead so you can inspect recent model outputs without loading full sessions. Use it to audit, compare, or continue conversations across multiple providers in real time.
How this skill works
Invoke the pend command with the provider name and an optional count to fetch the most recent replies. The tool reads persisted conversation context and returns the latest N replies for that provider, formatted for terminal display. It uses lightweight access to stored interactions so response retrieval is fast and keeps token usage minimal.
When to use it
- Quickly check the most recent response from a specific AI provider without opening the full application.
- Compare latest outputs across providers during multi-AI experimentation.
- Audit recent model behavior or verify the last message before continuing a session.
- Pull a small set of recent replies for debugging or logging purposes.
- Confirm what the user or an agent last saw before issuing follow-up instructions.
Best practices
- Specify the provider name exactly: gemini, codex, opencode, droid, or claude.
- Add a numeric argument to retrieve multiple recent replies (e.g., pend codex 3).
- Use in terminal workflows or scripts to keep integrations lightweight and automated.
- Run pend before issuing follow-ups to avoid duplicating or contradicting recent outputs.
- Keep persisted context compact to ensure fast retrieval and lower token costs.
Example use cases
- pend gemini — view the latest Gemini reply before deciding the next prompt.
- pend codex 3 — fetch the three most recent Codex responses while debugging code generation.
- pend claude — confirm the last Claude assistant reply before resuming a conversation.
- Integrate pend into CI scripts to log latest AI outputs during automated tests.
- Use pend opencode to quickly verify the last OpenCode transformation in a pipeline.
FAQ
Pass the provider name (gemini, codex, opencode, droid, claude) and optionally a number N to return the latest N replies.
How do I run it?
Execute pend with arguments in your terminal, for example: pend codex 2.