Hermes Loop Master
Impone specifiche, evidenze RED/GREEN e revisione avversaria del diff prima del rilascio del codice
Promosso
Cosa fa
Un coding-discipline loop che fa sì che l'agente scriva un checkable execution contract, implementi one slice at a time, provi behavior con RED/GREEN plus positive, negative, preservation and failure paths, quindi riveda il proprio diff against ten fake-done patterns prima di dichiarare work finished. Si attiva su multi-step, production-facing, security-sensitive o cross-session coding tasks. Include un executable harness che valuta gli loop artifacts, so the process is machine-checked rather than self-reported.
Rapporto di test
Ho corretto lo same duration-parser bug twice on identical repos, once free-hand and once under the loop. Both arms produced a green test suite, but the skill's mandated failure-path check exposed that the obvious regex fix silently returns 0 for 'banana' and '', 3600 for the typo '1h30', and 300 for '-5m' — a probe of ten inputs against both final versions diverged on five, with the no-skill version returning a wrong number where the skill version raises. The bundled harness_check.py è reale e funziona: it scored the loop artifacts 32/34, flagged two unchecked Done When boxes I had actually left open, correctly passed the good-loop and critical-loop fixtures and rejected the bad-loop fixture, and the repo's 36 unit tests pass. Costa three extra artifact files and roughly double the tool calls, so it earns its keep on risky work, not on typos. Note it is branded for Nous Research's Hermes Agent and its installer defaults to ~/.hermes, so Claude Code users must pass --target as in the command above.
Testato il: 2026-07-30 · Claude Code 2.x (agent harness)
Installazione
git clone https://github.com/s0heyl/hermes-loop-master.git mkdir -p ~/.claude/skills cd hermes-loop-master && bash install.sh --target ~/.claude/skills/hermes-loop-master
Comandi e prompt di esempio
/hermes-loop-masterImpone specifiche, evidenze RED/GREEN e revisione avversaria del diff prima del rilascio del codice
Gli skill si attivano con richieste in linguaggio naturale, senza comandi da ricordare. Dopo l'installazione, prompt come questi lo attivano (in inglese):
Fix this auth bypass in production, carefully, with proof it actually worksRefactor the payment webhook handler, this will span a few sessionsContinue yesterday's migration work under the RED/GREEN evidence loop