copilot-expert

Use cuando reportes/investigues bug del copilot module: loops infinitos, cards faltantes, conversaciones vacías al refresh, extracción que falla, trazas que mienten, tools que no ejecutan, costos disparados, routing equivocado, prompt cache roto, mutaciones que no persisten, channel format ignorado, voseo en outputs. Activa también para extender (agregar provider, tool, workflow, channel, KB chunk) sin romper invariantes. Triggers: 'bug copilot', 'loop copilot', 'card no aparece', 'conversación vacía', 'documento no extrae', 'agente no responde', 'tool no ejecuta', 'trazas mienten', 'tokens burned', 'extract_document_to_fields falla', 'propose_field_updates no persiste', 'format_for_channel ignored', 'voseo en respuesta', 'cache hit rate', 'cuántos tokens', 'agregar provider', 'agregar workflow', 'agregar tool transversal', 'agregar canal'.

Quality

98%

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

96%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.

The body is a high-quality, token-efficient operational manual: concrete, executable, and rigorously sequenced with validation checkpoints and feedback loops. Progressive disclosure is good with verified one-level-deep references, though some inline tables could be further externalized to tighten the overview.

Suggestions

Consider moving the large phase-by-phase map (F0-F11 table) and the full anti-pattern checklist table into a reference file, leaving only a pointer and the most critical anchors in SKILL.md to sharpen it as an overview.

The Budget Gating / plan-defaults table is detailed operational data that may belong in references/ alongside copilot-observability.md rather than inline, keeping the SSoT pointer as the anchor.

Dimension	Reasoning	Score
Conciseness	Extremely lean and dense; assumes Claude's competence throughout with zero explanation of basic concepts. Every table, anchor, and command earns its place (e.g., 'Regla cero: verificá que algo existe (grep + schema + registry + repo) ANTES de declarar').	5 / 5
Actionability	Fully executable guidance throughout: exact file paths, concrete SQL queries, specific test names (test_copilot_anchors.py:96), env vars (UPDATE_GOLDEN=1, RUN_LLM_JUDGE=1), and copy-paste bash/pytest commands covering the common cases.	5 / 5
Workflow Clarity	Multi-step processes are strictly sequenced with explicit validation checkpoints and feedback loops: 'Diagnóstico — orden estricto (no saltarse)', the 9-step Bug-fix protocol with 'RED → GREEN + quality gates', and a closing checklist with verification gates before code is touched.	5 / 5
Progressive Disclosure	Good structure with clear section headers and well-signaled one-level-deep references to real bundle files (references/copilot-resilience.md, references/copilot-observability.md, both confirmed present). It is not a 5 because the SKILL.md itself is fairly large with several dense inline tables (phase map, anti-pattern checklist, plan defaults) that could arguably live in separate reference files.	4 / 5
	Total	19 / 20 Passed

Description

100%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.

The description is exemplary: it concisely states concrete capabilities, provides comprehensive natural and technical trigger phrases, and explicitly answers both what the skill does and when to use it. It is tightly scoped to the copilot module, minimizing conflict risk.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions spanning both debugging ('loops infinitos, cards faltantes, conversaciones vacías al refresh, extracción que falla') and extending ('agregar provider, tool, workflow, channel, KB chunk'), giving comprehensive coverage of the skill's capabilities.	5 / 5
Completeness	Explicitly answers both 'what' (report/investigate copilot bugs, extend the module without breaking invariants) and 'when' via an explicit 'Use cuando...' clause followed by a dedicated 'Triggers:' list with concrete trigger phrases.	5 / 5
Trigger Term Quality	Extensive natural trigger phrases users would actually say ('bug copilot', 'card no aparece', 'conversación vacía', 'tokens burned', 'cache hit rate') plus technical synonyms and function names ('extract_document_to_fields falla', 'format_for_channel ignored'), covering both casual and technical phrasings.	5 / 5
Distinctiveness Conflict Risk	A clear niche — the copilot module — with highly specific triggers including function names and symptom descriptions, making conflict with unrelated skills minimal.	5 / 5
	Total	20 / 20 Passed

Validation

93%

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 — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
referenced_paths_exist	Referenced path issues: 2 missing, 1 deeper-than-1-level	Warning

	Total	15 / 16 Passed

Repository: alpacapurpura/ap_sales_agent
Path: .claude/skills/copilot-expert/SKILL.md
Commit: d31f7bc

Reviewed: ~19 hours 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.