CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

memory-config

Diagnose and configure MemSearch memory behavior. Use when the user asks about MemSearch configuration, plugin summarization, PROJECT.md/USER.md maintenance, memory directories, index health, provider routing, prompt files, or migration/compatibility questions.

Quality

85%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Critical

Do not install without reviewing

The canonical home for this skill is memory-config in zilliztech/memsearch

SKILL.md
Quality
Evals
Security

Quality

Content

71%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 content is a well-structured, highly actionable configuration guide with exemplary progressive disclosure — shared logic inline, platform detail split into six real reference files loaded conditionally. Its main weaknesses are conciseness (repeated disclaimers/pointers and version-sensitive details in the main flow) and a missing concrete `memsearch config set` example for the skill's central operation.

Suggestions

Add a concrete `memsearch config set` example with a real key, value, and scope (e.g., `memsearch config set --global plugins.claude-code.summarize.enabled true`) so the skill's primary operation is copy-paste executable.

Move version- and model-sensitive details (v0.4.11 project-config restrictions, default model names like claude-sonnet-4-6 / gpt-5-mini / gemini-3-flash-preview) into a clearly labeled compatibility section or the platform reference files to keep the main flow evergreen.

State the MemSearch-vs-host-identity disclaimer once and consolidate the repeated "see your platform reference file" pointers into the platform-detection section to trim redundant tokens.

DimensionReasoningScore

Conciseness

The body is dense and product-specific rather than padded with concepts Claude already knows, but it includes avoidable redundancy and time-sensitive detail: the MemSearch-vs-host identity disclaimer is stated twice (opening and again in 'Applying changes'), "see your platform reference file" is repeated about five times, and version/model specifics ("Since v0.4.11", "gpt-5-mini", "claude-sonnet-4-6", "gemini-3-flash-preview") sit in the main flow rather than a compatibility/deprecated section, which the guidelines penalize. This fits 'mostly efficient but could be tightened' better than level 4's 'minor instances'.

3 / 5

Actionability

The skill provides many copy-paste-ready commands (memsearch config list --resolved, uv tool install -U "memsearch[onnx]", the curl PyPI version check, index-state inspection) and concrete TOML key examples. It falls short of level 5 because the core action, `memsearch config set`, never gets a concrete example invocation with a key, value, and scope flag — the reader must infer the syntax for the skill's most important operation.

4 / 5

Workflow Clarity

The flow is clear and well-ordered: intent routing → diagnose first with explicit commands → scope choice (global vs project allowlist) → apply with `config set` → verify ("show the command, the resolved value, and whether a new session is needed"), plus error-recovery guidance for degraded index and failed maintenance states. It misses level 5 because the sequence is implied by section order rather than an explicit stepwise procedure, and some verification points remain implicit.

4 / 5

Progressive Disclosure

The body keeps shared logic inline and externalizes all platform-specific detail to six real, one-level-deep reference files (verified: references/claude-code.md, codex.md, dsh.md, openclaw.md, opencode.md, workbuddy.md exist and contain no nested references), with explicit conditional loading ("Read ONLY the one file matching your current environment") and environment markers for platform detection. This matches the level-5 anchor: clear overview, well-signaled references, easy navigation.

5 / 5

Total

16

/

20

Passed

Description

91%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 strong: it states a clear, concrete capability (diagnose and configure MemSearch memory behavior) and pairs it with an explicit, comprehensive 'Use when' trigger list covering all major feature areas. Its only weakness is that a few generic memory-related trigger terms could collide with a host platform's built-in memory configuration.

DimensionReasoningScore

Specificity

"Diagnose and configure MemSearch memory behavior" names two concrete actions and the when-clause enumerates specific scope areas (plugin summarization, PROJECT.md/USER.md maintenance, index health, provider routing, prompt files, migration/compatibility). It lists several specific capabilities with only minor gaps, fitting the level-4 anchor better than level 3 (which covers just 1-2 actions without comprehensive scope) and falling short of level 5's multiple distinct concrete actions.

4 / 5

Completeness

"Diagnose and configure MemSearch memory behavior" clearly answers what the skill does, and "Use when the user asks about..." explicitly answers when with concrete trigger phrases. This directly matches the level-5 anchor (both what and when with concrete triggers) rather than level 4, where the when-clause would be less explicit.

5 / 5

Trigger Term Quality

The trigger list "MemSearch configuration, plugin summarization, PROJECT.md/USER.md maintenance, memory directories, index health, provider routing, prompt files, or migration/compatibility questions" comprehensively covers the natural phrases a user would say, including concrete file names (PROJECT.md/USER.md). It matches the level-5 anchor (comprehensive natural terms including synonyms/extensions); level 4 would imply missing common variations, and none are evident for this product's niche.

5 / 5

Distinctiveness Conflict Risk

The MemSearch branding and product-specific triggers (PROJECT.md/USER.md, index health, provider routing) create a clear niche with minimal conflict risk. However, generic phrases like "memory directories" and "memory behavior" could overlap with a host agent's built-in memory/config system — a confusion the skill body itself acknowledges — so it is not fully level 5.

4 / 5

Total

18

/

20

Passed

Validation

100%

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

zilliztech/memsearch

Path

plugins/codex/skills/memory-config/SKILL.md

Reviewed

Table of Contents

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.