permissions-debug

Troubleshoot Kode permission prompts/denials (tool allowlists, commandAllowedTools, dontAsk fail-closed, subagent inheritance). Use when tools are unexpectedly blocked, permission prompts repeat, or behavior differs between main agent and subagents.

Quality

89%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

86%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A tight, well-structured debugging skill that respects the context window and sequences its triage logically with a verification loop. The main improvement room is converting a few high-level directions into concrete, runnable steps and tightening the verification loop into an explicit validate-fix-retry checkpoint.

Suggestions

Make the forensics step executable: give the exact command or glob to read session artifacts (e.g. 'tail the latest ~/.kode/<session>/messages.jsonl') rather than 'Inspect the latest session artifacts'.

Tighten the verification loop into an explicit validate-fix-retry cycle: 'Re-run the blocked action -> if still blocked, apply one minimal change -> repeat' with a clear stop condition.

For step 1 ('Look for the exact tool name and the rejection message'), specify where to find the rejection message (terminal output vs. ~/.kode/ error log) so Claude knows exactly where to look.

Dimension	Reasoning	Score
Conciseness	Lean and efficient: short labeled sections with no padding or explanation of concepts Claude already knows; every line earns its place.	5 / 5
Actionability	Provides concrete commands (/approved-tools list, /approved-tools remove <tool>) and concrete filesystem paths (~/.kode/, ~/.kode/**/tasks/), but some triage steps ('Look for the exact tool name', 'Inspect the latest session artifacts') remain directions rather than fully runnable instructions.	4 / 5
Workflow Clarity	Clear sequenced flow (fast triage -> verification loop -> forensics) with an explicit re-run-and-confirm validation loop; falls short of 5 because the verification loop is described abstractly rather than as a strict checkpoint with error-recovery branching.	4 / 5
Progressive Disclosure	Simple skill under 50 lines with no external references needed; content is well organized into clearly labeled sections, satisfying the simple-skill exception.	5 / 5
	Total	18 / 20 Passed

Description

92%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong, specific description that names concrete capabilities and provides explicit 'Use when...' triggers tied to recognizable user complaints. The only soft spot is trigger-term coverage, which is good but relies on Kode-specific jargon and omits common synonyms a user might utter.

Dimension	Reasoning	Score
Specificity	Lists multiple specific concrete actions across the permission system (tool allowlists, commandAllowedTools, dontAsk fail-closed, subagent inheritance) with comprehensive coverage; not merely a single named action.	5 / 5
Completeness	Explicitly answers both 'what' (troubleshoot Kode permission prompts/denials) and 'when' via a concrete 'Use when...' clause with multiple trigger conditions.	5 / 5
Trigger Term Quality	Includes natural phrases a user would say ('tools are unexpectedly blocked', 'permission prompts repeat', 'behavior differs between main agent and subagents'), but the trigger set leans on product-specific jargon and misses common synonyms like 'permission denied' or 'tool not allowed'.	4 / 5
Distinctiveness Conflict Risk	Occupies a clear niche (Kode permission debugging) with triggers that would not naturally match other skills, giving minimal conflict risk.	5 / 5
	Total	19 / 20 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: shareAI-lab/Kode-CLI
Path: packages/builtin-skills/skills/permissions-debug/SKILL.md
Commit: c7f6fcc

Reviewed: 3 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.