CtrlK
BlogDocsLog inGet started
Tessl Logo

llm-runtime-architecture

Use when designing how one canonical agent core runs across Codex, Claude Code, Gemini CLI, Cursor, and AGENTS.md-compatible tools.

57

Quality

64%

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

Fix and improve this skill with Tessl

tessl review fix ./.agents/skills/llm-runtime-architecture/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

72%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is exceptionally concise and well-structured into a procedure and output schema, but the steps read as production directives rather than fully executable guidance, and validation checkpoints remain implicit rather than explicit.

Suggestions

Add a concrete populated example of the runtime matrix (one row per runtime with sample values) so the expected output is unambiguous and copy-paste ready.

Promote validation to an explicit step — e.g. "6. Verify: confirm each runtime's verification command runs and unsupported capabilities are stated" — with a fix-and-retry loop.

Show one example adapter structure or a snippet of `.agentlas/global-commands.json` to make the adapter-writing steps actionable.

DimensionReasoningScore

Conciseness

At roughly 20 lines it is lean and efficient with no padding and no explanation of concepts Claude already knows — quoting the procedure "Keep `AGENTS.md` as the canonical behavior contract" and the field list in step 2, every token earns its place.

3 / 3

Actionability

It gives concrete guidance — a named file path ("Write or repair `.agentlas/global-commands.json`") and an explicit output schema ("runtime", "entry_point", "global_command", "adapter_files", "memory_access", "limitations", "verification") — but the steps are directives to produce rather than executable commands, and no populated matrix example or sample verification command is provided.

2 / 3

Workflow Clarity

The five numbered steps form a clear sequence and the per-runtime "verification" field plus step 5 ("State unsupported capabilities explicitly") imply checkpoints, but validation is never an explicit step and there is no validate→fix→retry feedback loop, leaving checkpoints implicit.

2 / 3

Progressive Disclosure

Under 50 lines with no need for external references, it is well-organized into clearly labeled "Procedure" and "Output" sections, which per the simple-skills scoring note earns a top score for progressive disclosure.

3 / 3

Total

10

/

12

Passed

Description

57%

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 has a clear, distinctive niche with explicit trigger guidance and strong product-name keywords, but it names only one action and leaves the concrete deliverables implicit, capping specificity and completeness.

Suggestions

Add a second concrete action and state the deliverable explicitly, e.g. "Produces a runtime matrix naming each runtime's entry point, adapter files, and verification command. Use when designing...".

Reword the trigger to match natural phrasing, e.g. "Use when you need an agent to run across Codex, Claude Code, Gemini CLI, and Cursor" to capture how users actually ask.

Surface common variations of the request ("port an agent to multiple CLIs", "AGENTS.md-compatible runtimes") as trigger keywords.

DimensionReasoningScore

Specificity

Quoting "designing how one canonical agent core runs across Codex, Claude Code, Gemini CLI, Cursor, and AGENTS.md-compatible tools" — it names a concrete domain and scope with several real runtimes, but offers only a single action verb ("designing"), not the multiple concrete actions the score-3 anchor requires.

2 / 3

Completeness

It opens with an explicit "Use when designing..." trigger clause, satisfying the when; however the what is only loosely embedded in that clause ("designing how one canonical agent core runs") with no separate statement of concrete deliverables, so it does not clearly answer both what AND when at the score-3 level.

2 / 3

Trigger Term Quality

It surfaces real product names users mention (Codex, Claude Code, Gemini CLI, Cursor, AGENTS.md), but the phrasing "designing how one canonical agent core runs across" is technical/awkward rather than a natural user utterance, missing common variations of how the request would actually be voiced.

2 / 3

Distinctiveness Conflict Risk

Quoting "one canonical agent core runs across Codex, Claude Code, Gemini CLI, Cursor, and AGENTS.md-compatible tools" carves out a precise niche anchored on named products, making it unlikely to trigger for the wrong skill.

3 / 3

Total

9

/

12

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.

Validation16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
agentlas-ai/Agentlas-OS
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.