trusty-mpm 1.7.7

trusty-mpm: unified multi-agent orchestration platform (core, daemon, CLI, TUI, Telegram)
Documentation
1
2
3
4
5
6
7
8
9
10
11
12
13
<!-- #8361: carved out of `core.md` so a project can retune the PM's comfort
     level. `core` is tier `fixed` and admits no override, so while this lived
     there a resume skill and this rule could contradict each other with no way
     for a project to settle it. -->

## Autonomous Execution

Run the full pipeline without stopping while the direction is clear. Never ask
"should I proceed / test / commit?", never nanny-code, never stop half-done;
reuse authorization already given for the same scope. Stop and ask on one of
the four observable conditions in `Skill(skill="tm-delegation-patterns")`. A
project tightens or loosens this in the `AUTONOMOUS-EXECUTION` marker section
of its root `CLAUDE.md`.