The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
fix: address final lint blockers for agent self-evaluation
Hawthorn149be89d397e
diff --git a/agents/agent-evaluator.md b/agents/agent-evaluator.mdindex 043171182..c44242ba2 100644--- a/agents/agent-evaluator.md+++ b/agents/agent-evaluator.md@@ -203,4 +203,4 @@ TOP IMPROVEMENTS:3. [Completeness] Handle 429, connection errors, and timeoutVERDICT: Redo with specific fixes. Weakest axis: Accuracy (2/5).-```\ No newline at end of file+```diff --git a/skills/agent-self-evaluation/SKILL.md b/skills/agent-self-evaluation/SKILL.mdindex 0e1a2fd67..4e241380a 100644--- a/skills/agent-self-evaluation/SKILL.md+++ b/skills/agent-self-evaluation/SKILL.md@@ -85,7 +85,7 @@ If any axis scored 3 or below:1. State what you would do differently2. If the gap is fixable in < 30 seconds (missing link, unclear phrasing), fix it now-3. If the gap requires rework, flag it explicitly: "This axis scored __ because __. Re-running with __ would likely raise it to __."+3. If the gap requires rework, flag it explicitly: "This axis scored [reason] because [evidence]. Re-running with [specific fix] would likely raise it to [score]."## Code Examples@@ -134,7 +134,7 @@ Overall: 2.8 — Wrong library used. Needs httpx rewrite.### "Everything is a 5"```-❌ Accuracy: 5 — All good.+FAIL: Accuracy: 5 — All good.Completeness: 5 — Everything covered.Clarity: 5 — Clear.```@@ -144,7 +144,7 @@ No evidence cited. This is self-congratulation, not evaluation. A real 5 require### Over-penalizing for scope creep```-❌ Completeness: 2 — Didn't handle WebSocket connections or+FAIL: Completeness: 2 — Didn't handle WebSocket connections orgRPC streaming (user didn't ask for these)```@@ -153,7 +153,7 @@ Only evaluate against what the user actually requested, not what you could have### Using the evaluation to re-litigate```-❌ "As I said earlier, this approach is wrong. Score: 1"+FAIL: "As I said earlier, this approach is wrong. Score: 1"```The evaluation is about the delivered output, not about re-arguing design decisions that were already made. If the approach was wrong, that should have been caught before delivery.@@ -161,7 +161,7 @@ The evaluation is about the delivered output, not about re-arguing design decisi### Mixing personal preference with objective gaps```-❌ "Score: 3. I don't like Python decorators."+FAIL: "Score: 3. I don't like Python decorators."```"Don't like" is not evidence. Cite a concrete readability, testability, or correctness concern, or leave the score at 4+.diff --git a/skills/agent-self-evaluation/references/evaluation-criteria.md b/skills/agent-self-evaluation/references/evaluation-criteria.mdindex 9a352bf1a..fbb3cf90a 100644--- a/skills/agent-self-evaluation/references/evaluation-criteria.md+++ b/skills/agent-self-evaluation/references/evaluation-criteria.md@@ -56,7 +56,7 @@ This reference provides concrete scoring anchors for each axis. Use it when you'### When the user gave unclear instructions-If the user's request was ambiguous, do NOT penalize completeness for not reading minds. Instead, note in the evaluation: "User's request was ambiguous about __. I chose interpretation __. If they meant __, this score would drop to __."+If the user's request was ambiguous, do NOT penalize completeness for not reading minds. Instead, note in the evaluation: "User's request was ambiguous about [scope]. I chose interpretation [chosen interpretation]. If they meant [alternative interpretation], this score would drop to [score]."### When the task is inherently simplediff --git a/skills/agent-self-evaluation/templates/evaluation-report.md b/skills/agent-self-evaluation/templates/evaluation-report.mdindex 467370922..bbc06d4ba 100644--- a/skills/agent-self-evaluation/templates/evaluation-report.md+++ b/skills/agent-self-evaluation/templates/evaluation-report.md@@ -83,4 +83,4 @@ Skip the evaluation if:| ≥4.5 | Deliver as-is. No changes needed. || 3.5–4.4 | Flag top improvement but deliver. Fix if <30 seconds. || 2.5–3.4 | State what you'd change. Ask user: "Should I redo [axis] or deliver as-is?" |-| <2.5 | Don't deliver. Say: "This scored __ because __. Let me redo this with [specific fix]." Then redo. |+| <2.5 | Don't deliver. Say: "This scored [score] because [evidence]. Let me redo this with [specific fix]." Then redo. |