What it does. This package targets Prime Agent 0.7.2. Before recursive use, set
rlmMaxDepth
to
4
in Prime Agent settings and start a fresh session so this Python-backed skill and its extension are loaded.
Start only for an explicit
/autoprompt
request. A bare invocation reports the recorded frontier and stops; resume requires an explicit
resume
request.
Highlights
- Concurrency:
tokensaver
, wide
, or custom max_subs=N
.
- Agent selection:
off
/inherit only. Confirm that every child inherits the already-selected parent model; no per-child model routing selector is available.
apply
: apply.mdbackend-build
: backend-build.mdbackend-fix
: backend-fix.md
When to use it. Run the Autoprompt orchestration loop on Prime Agent through a topology-enforcing native RLM dispatcher. Use only when the user explicitly invokes Autoprompt or asks to run the loop.
Install
npx skills add Spielewoy/autoprompt-skill --skill autoprompt
Or copy the skill folder manually:
git clone https://github.com/Spielewoy/autoprompt-skill.git
cp -r autoprompt-skill/agents/prime/skills/autoprompt/ ~/.claude/skills/autoprompt
Source: Spielewoy/autoprompt-skill — 972 stars, MIT, last pushed 2026-08-30.
