CtrlK
BlogDocsLog inGet started
Tessl Logo

agent-memory-discipline

Teaches when to recall from long-term memory before acting and when to save durable decisions, corrections and failures afterwards. Use when a memory tool or MCP memory server is connected but the agent is not using it consistently, when the user complains that the assistant forgets preferences, conventions or past decisions between sessions, or when setting up persistent memory for a project. Works with any memory backend: a folder of Markdown files, a local MCP server, or a managed service.

68

Quality

83%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

78%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 well-structured, mostly lean instruction-only skill with concrete rules, a worked example, and a closing checklist; the guidance is actionable despite the absence of code. The main gaps are minor editorializing and the lack of an explicit validation/feedback checkpoint in the recall-save workflow.

Suggestions

Trim rhetorical flourishes (e.g. 'the single most destructive habit in agent memory, and it is invisible until someone asks a question about old code') to tighten conciseness without losing the rule.

Add a concrete recall/save template — e.g. a search-query construction rule and a one-line memory-entry format like 'FACT (when: date range, source: file/commit)' — to lift actionability beyond abstract lists.

Turn the recall→save→close flow into an explicit numbered sequence with a validation checkpoint, such as 'after recalling, confirm the entry's validity window still applies before acting on it', to add a feedback loop.

DimensionReasoningScore

Conciseness

Mostly lean rule-based prose with no explaining of concepts Claude already knows, but occasional editorializing flourishes ('the single most destructive habit in agent memory', 'Paraphrase drifts') could be trimmed; sits above the verbose midpoint but short of every-token-earns-its-place.

4 / 5

Actionability

Provides concrete lists (when to recall, when to save, what fields each entry carries) and a worked example with a sample memory entry, but stays backend-agnostic so there is no exact tool-call syntax or template to copy; mostly actionable with minor gaps.

4 / 5

Workflow Clarity

Clear recall-before / save-after structure plus a closing checklist and a 4-step worked example, but there is no explicit validation checkpoint (e.g. confirm an entry's validity window before acting on it); minor validation gap keeps it below 5.

4 / 5

Progressive Disclosure

Self-contained SKILL.md with no bundle files and well-organized section headers (Recall, Save, Write it so it survives, Close don't overwrite, etc.), a worked example, and a checklist; navigation is easy and there are no nested references.

5 / 5

Total

17

/

20

Passed

Description

87%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 description that clearly states what the skill does and gives multiple concrete, natural-language trigger conditions. It is third-person throughout with no over-claims; the only room for improvement is slightly more granular action wording and a few more synonyms.

DimensionReasoningScore

Specificity

Names concrete actions — 'recall from long-term memory before acting' and 'save durable decisions, corrections and failures afterwards' — with enumerated save targets; stops short of the granular action lists of a 5 because the two meta-actions stay somewhat abstract.

4 / 5

Completeness

Explicitly answers both 'what' (teaches when to recall and when to save) and 'when' with three concrete trigger clauses, matching the anchor for clearly and explicitly answering both with concrete trigger phrases.

5 / 5

Trigger Term Quality

Includes natural user phrasings like 'the assistant forgets preferences, conventions or past decisions between sessions' and 'setting up persistent memory for a project'; good coverage but a few common synonyms (e.g. 'remember', 'keep track of') are absent.

4 / 5

Distinctiveness Conflict Risk

Occupies a clear niche — memory-tool-not-being-used-consistently / forgetting-between-sessions — with triggers specific enough to avoid firing for unrelated skills; minimal conflict risk.

5 / 5

Total

18

/

20

Passed

Validation

93%

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

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository
davepoon/buildwithclaude
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.