autoresearch-loop
Use this when running an automated performance-experiment loop on a repo: one change at a time, kept only if the correctness gate passes and the benchmark gets faster.
Grok Bots DirectorySubmit a Grok Botby Kevin
Last checked Template updated
Makes a code repo faster by running one experiment at a time, keeping only the changes that pass your correctness check and beat your benchmark. It finishes with a single reviewable draft PR or set of local commits, plus a short before-and-after report.
Use this when running an automated performance-experiment loop on a repo: one change at a time, kept only if the correctness gate passes and the benchmark gets faster.
Use this when starting any non-trivial task that might burn LLM or agent quota (long coding runs, experiment loops, long research, bulk clones). Estimate cost against remaining weekly quota and stop for approval if it would exceed 10%.
Use this for the first conversation with a new owner of the Autoresearch bot, to collect the repo, benchmark, correctness gate, and caps before the first run.
Leapfrog is a job desk: confidence-scored routing across Claude Code, Codex, and OpenCode, proof-before-done with real changed files, one-writer locks, and usa…
by jacky wine
Dual-layer digital garden: flomo as L1 (flash notes, tweet ingest) and local Obsidian as disk (full-text clippings, dense wikilinks). Curates and cross-links;…
A patient Minecraft coach for people who play with a Hermes bot on a Paper + Geyser server. It teaches the in-game phrases and checks that the bot really is fo…
by headyrobot
A live score tracker that pings you on every score, quarter or half end, overtime, and the final for the games you pick. It can follow one game or your team's…
by Jonathan
Finds what's really draining your Android battery, even when the battery usage list in Settings doesn't show anything unusual. Runs a full scan using ADB and s…