بنقرة واحدة
speckit-memorylint-load-agents
Spec-kit workflow command: speckit-memorylint-load-agents
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
القائمة
Spec-kit workflow command: speckit-memorylint-load-agents
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
Use when a GitHub Pull Request has unresolved review threads, pending reviews, stale/outdated threads, local findings ingestion, or needs mandatory final-gate proof in one PR-scoped session.
Optional after-specify refinement gate. Bridges an installed obra/superpowers brainstorming skill into the active Spec Kit spec.md without creating a second design document or replacing speckit.specify.
Diagnose the five Superpowers disciplines used by Superb and report the focused hook and standalone-command contract.
Read-only implementation critique against active Spec Kit artifacts.
Apply systematic root-cause investigation to the current failing task.
Complete branch handoff after Spec Kit convergence and fresh checks.
| name | speckit-memorylint-load-agents |
| description | Spec-kit workflow command: speckit-memorylint-load-agents |
| compatibility | Requires spec-kit project structure with .specify/ directory |
| metadata | {"author":"github-spec-kit","source":"memorylint:commands/load-agents.md"} |
$ARGUMENTS
You are the Core Rule Enforcer for the current workspace.
Load the root AGENTS.md before the planning phase, fail fast if it is missing,
and emit structured proof of what was loaded so downstream planning can inherit
the same rules.
AGENTS.md from the workspace root.AGENTS.md is missing, unreadable, or
cannot be loaded, STOP immediately. Do not begin planning, do not generate or
update plan.md or tasks.md, and do not continue to any subsequent step.On success, output a short confirmation plus a machine-readable JSON payload with at least:
{
"workspace_root": "/path/to/workspace",
"agents_path": "AGENTS.md",
"agents_sha256": "<sha256>",
"rule_count": 0,
"sections": [],
"rule_summaries": [
{
"rule_id": "R-001",
"line_range": "10",
"category": "workflow",
"summary": "Use focused commits with Conventional Commit style."
}
]
}
This output is the verifiable before_plan gate record. It must prove:
If loading fails, output a clear failure and stop immediately:
ERROR: Mandatory before_plan gate failed: could not load AGENTS.md from the workspace root. Planning cannot proceed. Remediation: ensure AGENTS.md exists, is readable, and that the workspace root is correct, then retry.