CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

agentmemory-architecture

How agentmemory is built, the iii engine primitives it runs on, its storage model, ports, and the viewer. Use when reasoning about how memory is stored or retrieved end to end, when extending the system, or when answering how agentmemory works under the hood.

Quality

90%

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

92%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 a lean, dense architecture reference that respects the token budget while surfacing exact flags, ports, and defaults Claude could not otherwise know. It is well-organized for navigation and reasoning; the only gap is the absence of complete executable command examples.

DimensionReasoningScore

Conciseness

Every sentence carries system-specific information Claude cannot know — exact environment flags ("EMBEDDING_PROVIDER=local", "AGENTMEMORY_ALLOW_AGENT_SDK=true"), ports ("3111", "3113", "49134"), defaults ("file-backed KV store by default... it is not SQLite"), and useful negative clarifications. There is no padding and no explanation of concepts Claude already knows, so every token earns its place.

5 / 5

Actionability

The body provides concrete, copy-paste-ready specifics — exact env var names and values, CLI flags ("--port <N> or --instance <N>"), and the viewer URL — which is actionable guidance for an instruction-only reference skill. It stops short of 5 because there are no complete command examples (e.g. an actual invocation selecting another port) and some sections are purely descriptive.

4 / 5

Workflow Clarity

This is a single-purpose, under-50-line reference skill with unambiguous purpose and clear topic-by-topic organization, so the simple-skill exception applies. No destructive or batch operations are involved, and extension guidance is even compressed into one concrete rule ("new capability is a new function plus a trigger").

5 / 5

Progressive Disclosure

The skill is self-contained with no bundle files, well-organized sections, and a clearly signaled "See also" list routing surfaces, hooks, and config details to sibling skills exactly one level deep. Per the under-50-lines guidance, well-organized sections alone merit the top score here.

5 / 5

Total

19

/

20

Passed

Description

83%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 covers and gives three explicit, natural trigger conditions in third person without fluff. Its main weaknesses are minor: a few missing natural synonyms and slight overlap risk with closely related sibling skills.

DimensionReasoningScore

Specificity

The description names the domain and lists several concrete coverage areas — "the iii engine primitives it runs on, its storage model, ports, and the viewer" — matching the several-specific-items-with-minor-gaps anchor. It falls short of 5 because the retrieval model and data boundary, covered prominently in the body, are not mentioned.

4 / 5

Completeness

It explicitly answers both: what ("How agentmemory is built, the iii engine primitives it runs on, its storage model, ports, and the viewer") and when, with three concrete trigger clauses ("Use when reasoning about how memory is stored or retrieved end to end, when extending the system, or when answering how agentmemory works under the hood"). This matches the anchor where both what and when are explicit with concrete trigger phrases, exceeding the level-4 anchor whose 'when' is less specific.

5 / 5

Trigger Term Quality

It includes natural phrases users would say such as "how memory is stored or retrieved", "extending the system", and "how agentmemory works under the hood", giving good keyword coverage. A few natural variations (e.g. "architecture", "internals", "debug agentmemory") are missing, so it does not reach comprehensive synonym coverage.

4 / 5

Distinctiveness Conflict Risk

The named-system focus and "under the hood" framing give it a clear niche with mostly distinct triggers. However, sibling skills in the same family (agentmemory-config, agentmemory-mcp-tools) create minor overlap risk — e.g. "extending the system" could also pull for the config skill — so it does not fully meet the minimal-conflict anchor.

4 / 5

Total

17

/

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

rohitg00/agentmemory

Path

plugin/skills/agentmemory-architecture/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.