agent-memory-systems
Memory is the cornerstone of intelligent agents. Without it, every interaction starts from zero. This skill covers the architecture of agent memory: short-term (context window), long-term (vector stores), and the cognitive architectures that organize them.
38
24%
Does it follow best practices?
Impact
Pending
No eval scenarios have been run
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./skills/antigravity-agent-memory-systems/SKILL.mdQuality
Discovery
22%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This description reads like an academic course module summary rather than a functional skill description. It lacks concrete actions Claude would perform, contains no explicit trigger guidance ('Use when...'), and relies heavily on abstract conceptual language rather than practical, user-facing terminology. The philosophical opening sentence ('Memory is the cornerstone of intelligent agents') wastes space that should be used for actionable detail.
Suggestions
Replace the abstract opening with concrete actions, e.g., 'Designs and implements memory systems for AI agents, including vector store setup, context window management, and retrieval-augmented generation (RAG) pipelines.'
Add an explicit 'Use when...' clause with natural trigger terms, e.g., 'Use when the user asks about agent memory, vector databases, RAG, embedding storage, long-term recall, or context management.'
Remove philosophical framing ('cornerstone of intelligent agents', 'every interaction starts from zero') and replace with specific, distinguishable capabilities that differentiate this skill from general AI architecture skills.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | The description uses abstract, conceptual language ('cornerstone of intelligent agents', 'cognitive architectures') rather than listing concrete actions Claude can perform. No specific actions like 'store', 'retrieve', 'search', or 'index' are mentioned. | 1 / 3 |
Completeness | The description vaguely addresses 'what' (covers architecture of agent memory) but provides no 'when' clause or explicit trigger guidance. It reads more like a course syllabus topic than a skill description, and the missing 'Use when...' clause would cap this at 2 regardless, but the 'what' is also weak. | 1 / 3 |
Trigger Term Quality | Contains some relevant technical terms like 'agent memory', 'context window', 'vector stores', and 'cognitive architectures', but these are more academic/jargon than natural user language. Missing common user phrases like 'remember', 'recall', 'store information', 'knowledge base'. | 2 / 3 |
Distinctiveness Conflict Risk | The focus on 'agent memory' and 'vector stores' provides some specificity, but the broad framing ('cornerstone of intelligent agents') and overlap with general AI architecture topics could cause conflicts with other AI/agent-related skills. | 2 / 3 |
Total | 6 / 12 Passed |
Implementation
27%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill is comprehensive in coverage but severely violates token efficiency by including everything in a single monolithic file. It explains many concepts Claude already understands, includes multiple redundant code examples (e.g., contextual chunking appears twice), and lacks the progressive disclosure structure needed for a skill of this scope. The actionability is moderate—code examples exist but rely on undefined helpers and unverified API surfaces.
Suggestions
Split into SKILL.md (overview + quick start + principles) with separate reference files: VECTOR_STORES.md, CHUNKING.md, SHARP_EDGES.md, PATTERNS.md — each linked from the main file
Remove explanations of concepts Claude already knows (what semantic/episodic/procedural memory are, what vector databases do, what chunking is) and replace with terse decision rules
Add explicit validation checkpoints to workflows involving destructive operations (memory deletion, consolidation, embedding migration) — e.g., 'verify count before/after deletion'
Ensure code examples have all necessary imports and replace undefined helper functions with actual implementations or explicit notes that they're project-specific stubs
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Extremely verbose at ~600+ lines. Explains concepts Claude already knows (what semantic/episodic/procedural memory are, what vector stores do, basic chunking concepts). Includes extensive decision matrices, multiple vector store examples, and lengthy sharp edges sections that could be dramatically condensed or split into reference files. | 1 / 3 |
Actionability | Contains many code examples that appear executable, but they use framework-specific APIs (LangMem) that may not match actual library interfaces, and several examples are wrapped in triple-quoted strings rather than proper code blocks. The code is illustrative rather than verified copy-paste ready, with missing imports and undefined helper functions (e.g., cluster_by_similarity, summarize, embed). | 2 / 3 |
Workflow Clarity | Individual patterns have clear steps, but there's no overarching workflow for 'how to design and implement an agent memory system.' The background memory formation and memory decay patterns have reasonable sequences, but lack explicit validation checkpoints. For destructive operations like memory deletion/archival, there are no verification steps. | 2 / 3 |
Progressive Disclosure | This is a monolithic wall of text with everything inline. The vector store examples, chunking strategies, sharp edges, and validation checks could all be split into separate reference files. There are no references to external files despite the content being far too long for a single SKILL.md overview. | 1 / 3 |
Total | 6 / 12 Passed |
Validation
81%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 9 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
skill_md_line_count | SKILL.md is long (1085 lines); consider splitting into references/ and linking | Warning |
frontmatter_unknown_keys | Unknown frontmatter key(s) found; consider removing or moving to metadata | Warning |
Total | 9 / 11 Passed | |
- Repository
- boisenoise/skills-collections
- Commit
636b862
- 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.