Name: memory-engineering
Rating: 76.8 (1 reviews)
Author: alirezarezvani

memory-engineering

Use when designing, reviewing, or paying for an agent memory system — adding memory to an agent, choosing between long-context / RAG / graph / agentic memory, auditing what a CLAUDE.md or memory directory actually holds, deciding what to keep and what to expire, or when a memory store keeps growing and nobody has said what leaves it. Prices the write path, picks which cost to pay, classifies records as facts / skills / logs, and refuses a design that has no forgetting policy.

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

The canonical home for this skill is memory-engineering in alirezarezvani/claude-skills

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 lean, highly actionable, and built around an explicitly gated five-step workflow with feedback loops, scoring top marks on conciseness, actionability, and workflow clarity. It loses a point on progressive disclosure because the referenced bundle files (references, scripts, assets) are not present in the bundle.

Suggestions

Ship the referenced bundle files under references/, scripts/, and assets/ so the one-level-deep navigation links in the 'References and assets' and 'Workflow' sections resolve.

Add a one-line note near the workflow stating where the scripts live if they are provided out-of-band, or remove the script invocations if the bundle is intended to be instruction-only.

Dimension	Reasoning	Score
Conciseness	Lean and efficient with no padding of concepts Claude already knows; the brief 'What this does' framing and the four-lenses table each earn their tokens; not the 4 anchor because there is no over-explanation to trim.	5 / 5
Actionability	Five copy-paste-ready 'python scripts/...' commands with exact flags and a scripts table giving exit-code meanings; covers common cases concretely; not the 4 anchor because guidance is fully executable rather than having minor gaps.	5 / 5
Workflow Clarity	A clearly sequenced five-step workflow with an explicit validation gate (Step 4, 'exit 4 is a stop') and a feedback loop in Step 5 ('prove each pass by hand before scheduling'); the destructive/batch cap-at-3 does not apply because validation is present; not the 4 anchor because checkpoints are explicit.	5 / 5
Progressive Disclosure	Clear overview with well-signaled one-level-deep references and a well-organized section structure, but the referenced bundle files (references/, scripts/, assets/) do not exist on disk, so the disclosure structure cannot be fully traversed; not the 5 anchor because some referenced navigation targets are missing, and not the 3 anchor because the on-page structure and signaling are clearly good.	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 third-person, lean, and explicitly pairs a 'Use when...' trigger list with concrete capabilities, scoring at the top of every dimension. It names the domain precisely and includes synonyms and file artifacts that a user would naturally say.

Dimension	Reasoning	Score
Specificity	Lists multiple specific concrete actions — 'Prices the write path, picks which cost to pay, classifies records as facts / skills / logs, and refuses a design that has no forgetting policy' — with comprehensive coverage; not the 4 anchor since there are no meaningful gaps.	5 / 5
Completeness	Explicitly answers both: a 'Use when...' clause with concrete trigger phrases for 'when', and the Prices/picks/classifies/refuses verbs for 'what'; not the 4 anchor because both halves are concrete and explicit.	5 / 5
Trigger Term Quality	Comprehensive natural-term coverage including synonyms ('long-context / RAG / graph / agentic memory') and concrete artifacts users name ('CLAUDE.md or memory directory', 'memory store keeps growing'); not the 4 anchor because nearly every natural phrasing is present.	5 / 5
Distinctiveness Conflict Risk	A clear niche (agent memory systems with a cost-and-forgetting framing) with distinct triggers; minimal overlap with adjacent skills; not the 4 anchor because the niche is sharply defined.	5 / 5
	Total	20 / 20 Passed

Validation

81%

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 — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
relative_links	Relative link issues: 7 missing	Warning
referenced_paths_exist	Referenced path issues: 23 missing	Warning

	Total	13 / 16 Passed

Repository: alirezarezvani/claude-skills
Path: .gemini/skills/memory-engineering/SKILL.md
Commit: 19392f7

Reviewed: ~7 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.