retrieval-reflex

When/what to retrieve — open the brain page for a salient entity before answering from memory.

Quality

78%

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

Fix and improve this skill with Tessl

tessl review fix ./recipes/retrieval-reflex/skills/retrieval-reflex/SKILL.md

Quality

Content

100%

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 lean, well-structured instruction skill: a clear when-to-retrieve policy, a concrete escalation/stop spec with named MCP tools, and a verify-before-asserting checkpoint. It assumes Claude's competence and avoids concept over-explanation while staying self-contained.

Dimension	Reasoning	Score
Conciseness	Lean and well-organized; the brief address-book analogy encodes the retrieval reflex without explaining concepts Claude already knows, and every section earns its place.	3 / 3
Actionability	Provides concrete, executable guidance — named MCP tools (`get_page <slug>`, `graph`, `backlinks`) and a precise three-level escalation spec with stop conditions — which is actionable for an instruction-only skill.	3 / 3
Workflow Clarity	The retrieval spec is a clearly numbered escalation sequence with explicit stop conditions and a verify-before-asserting checkpoint ('verify against the brain first'); as read-only retrieval it needs no destructive-feedback loop.	3 / 3
Progressive Disclosure	Under 50 lines with well-organized sections and clear one-level 'See also' pointers to sibling skills (query, brain-ops), satisfying the simple-skill allowance with no nested or multi-level references.	3 / 3
	Total	12 / 12 Passed

Description

57%

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 concisely conveys a distinct, well-scoped reflex with explicit what/when framing. Its weaknesses are jargon-heavy trigger language and an abstract rather than concrete when-condition, with the natural trigger terms residing in the frontmatter triggers field instead of the description.

Suggestions

Add a 'Use when...' clause with concrete natural triggers (e.g., 'Use when the user asks who someone is, what we know about a company, or to tell them about a person/project') so the description carries explicit trigger terms rather than relying on the separate triggers field.

Name more than one concrete capability in the description (e.g., 'open a brain page, resolve an unrecognized name, and pull linked neighbors before answering') to lift specificity.

Replace internal jargon like 'salient entity' and 'brain page' with user-facing phrasing in the trigger portion so the keywords match what users actually say.

Dimension	Reasoning	Score
Specificity	Names one concrete action ('open the brain page for a salient entity') and the retrieval domain, but lists only a single action rather than multiple concrete capabilities, matching the 'names domain and some actions' anchor rather than the multi-action level.	2 / 3
Completeness	It states both what ('open the brain page for a salient entity') and a when ('When/what to retrieve ... before answering from memory'), but the when is an abstract condition rather than concrete explicit triggers, so per the missing-'Use when' cap it stays at 2.	2 / 3
Trigger Term Quality	Contains relevant keywords like 'retrieve' and 'before answering from memory', but leans on internal jargon ('salient entity', 'brain page') rather than natural user phrasings, with the concrete natural triggers ('who is', 'tell me about') living only in the separate frontmatter triggers field.	2 / 3
Distinctiveness Conflict Risk	Carves a clear niche — on-demand retrieval of a salient entity's brain page before answering from memory — that is unlikely to trigger the wrong skill and is distinct from sibling query/brain-ops skills.	3 / 3
	Total	9 / 12 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
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: garrytan/gbrain
Path: recipes/retrieval-reflex/skills/retrieval-reflex/SKILL.md
Commit: c44cdb5

Reviewed: about 20 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.