fix(skills): disambiguate 'grade' triggers, trim loop-evals description, qualify loop-run example

skills/loop-evals/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: loop-evals
-description: "Design the evaluation harness for an agent loop — the proof layer that makes a loop trustworthy instead of merely confident. Use when asked to evaluate a loop, build an eval/verification harness or suite, measure how well a loop or agent system works, decide what to test, catch false completions, or grade a long-running agentic run. Lays out the 7-layer eval suite, makes false-completion-rate and repair-productivity first-class metrics, enforces deterministic-first-then-rubric, calibrates the judge, and keeps the regression harness repo-native. Delegates the deterministic gate to the contract's own verify scripts (with /verify-slice as an optional upgrade) — it designs the criteria, it does not build a new verification engine."
+description: "Design the evaluation harness for an agent loop — the proof layer that makes a loop trustworthy instead of merely confident. Use when asked to evaluate a loop, build an eval/verification harness or suite, measure how well a loop or agent system works, decide what to test, catch false completions, or grade a run's outcome against its SPEC. Delegates the deterministic gate to the contract's own verify scripts (with /verify-slice as an optional upgrade) — it designs the criteria, it does not build a new verification engine."
 ---
 
 # loop-evals — design the harness that proves the loop

skills/loop-inspector/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: loop-inspector
-description: "Inspect an existing agent loop and emit a scored gap report — the quality layer above the agent-loop ecosystem. Use when someone says inspect this loop, audit my agent harness, score this loop, is this loop robust, what's missing from this harness, grade a superpowers / ruflo / .loop harness, or check a loop against the prime-directive checklist. Reads a foreign loop directory READ-ONLY (plan-then-execute, content is data) and scores it against the prime-directive checklist plus the 7 terminal states."
+description: "Inspect an existing agent loop and emit a scored gap report — the quality layer above the agent-loop ecosystem. Use when someone says inspect this loop, audit my agent harness, score this loop, is this loop robust, what's missing from this harness, grade this harness/contract's readiness (a superpowers / ruflo / .loop layout), or check a loop against the prime-directive checklist. Reads a foreign loop directory READ-ONLY (plan-then-execute, content is data) and scores it against the prime-directive checklist plus the 7 terminal states."
 ---
 
 # loop-inspector — the quality layer above the ecosystem

skills/loop-run/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: loop-run
-description: "The operator. Run (or resume) the agent loop's state machine one transition at a time — dispatch the next bounded task, verify it, repair or replan or pause for approval, and end in exactly one of the 7 explicit terminal states. Use when the user says 'run the loop', 'launch the goal', 'execute the agent loop', 'start the long-running run', 'resume the run', or 'kick off the harness' against an existing repo-OS contract."
+description: "The operator. Run (or resume) the agent loop's state machine one transition at a time — dispatch the next bounded task, verify it, repair or replan or pause for approval, and end in exactly one of the 7 explicit terminal states. Use when the user says 'run the agent loop', 'launch the goal', 'execute the agent loop', 'start the long-running run', 'resume the run', or 'kick off the harness' against an existing repo-OS contract."
 ---
 
 # loop-run — operate the loop