Search: "propose"

Validates security patches with reproducible baseline-versus-patched evidence, including original exploits, root-cause variants, behavior preservation, regressions, and newly introduced security failures. Use after a patch exists and before accepting, merging, or reporting it as fixed; also use when an AI-generated patch, remediation commit, pull request, or proposed upstream fix needs adversarial post-patch validation across any language.

Analyzes user-requested Screenpipe history windows to detect repeated research workflows, match existing scientific skills, and stage new skill drafts or composition recipes for review. Requires a reachable Screenpipe HTTP API, normally on localhost:3030. Detection and embedding inference run locally; the selected LLM receives redacted app/title cluster summaries and matched skill descriptions. Use only when the user explicitly asks to analyze their recent work and propose skills.

You MUST use this before any creative work (features, products, content, strategy, systems, or behavior changes). Start by classifying what we’re brainstorming, then run thorough one-question-at-a-time discovery, propose 2–3 approaches, and converge on a validated plan/spec.

Categorize messages and propose replies and next steps.

Reproduce a bug report, judge severity, and propose the next step. Use when an issue, ticket, or error report needs a first response.

Use this skill to triage inbound customer and user feedback for affiliate-mcp in prepare-and-approve mode: read new and updated GitHub issues and the support inbox, draft grounded replies from the repo's own docs, propose triage labels, and escalate anything ambiguous, tense, or money-related to the operator. It drafts replies and proposes labels; it never posts a public comment or sends an email itself. Trigger on: "triage the support queue", "draft replies to new issues", "any new user feedback?", "handle customer service", "what did users report today?".

Operating procedure for monitors, alerts, scheduled briefs and reports in Adako, the MCP server and REST API that connects this assistant to Google Ads, Meta Ads, ChatGPT Ads, TikTok Ads and LinkedIn Ads. Use it whenever the user says "alert me when", "watch this", "let me know if", "send me a weekly summary", "email me a report", or asks what is being watched, why an alert fired, or how to pause a schedule. It covers the 13 monitoring and reporting tools, the metrics and operators a rule accepts, how to test a rule before saving it, how monitor-proposed actions reach the approval inbox, and the plan the tools need.

Operating procedure for monitors, alerts, scheduled briefs and reports in Adako, the MCP server and REST API that connects this assistant to Google Ads, Meta Ads, ChatGPT Ads, TikTok Ads and LinkedIn Ads. Use it whenever the user says "alert me when", "watch this", "let me know if", "send me a weekly summary", "email me a report", or asks what is being watched, why an alert fired, or how to pause a schedule. It covers the 13 monitoring and reporting tools, the metrics and operators a rule accepts, how to test a rule before saving it, how monitor-proposed actions reach the approval inbox, and the plan the tools need.

Introducing Lossless Agentic Compression. We introduce a method, and codebase, for losslessly compressing files using AI Agents, and explore the implications that come with this. Furthermore, we propose a definition and generalizable measure of intelligence, extending classic works.