CtrlK
BlogDocsLog inGet started
Tessl Logo

agent-system

Grida AI agent system work: `@grida/daemon` (DaemonServer, loopback HTTP perimeter, files/workspaces, secrets store, daemon discovery) and `@grida/agent` (the agent tenant: sessions, providers/BYOK, runtime/tool execution, skills discovery, prompts, tiers, sandbox hosts). Use for `packages/grida-daemon/**`, `packages/grida-ai-agent/**`, desktop sidecar protocol changes, agent chat transport, and bugs in agent state or streams. For pure Electron window, preload, menu, deep-link, or CDP work, use `desktop`.

68

Quality

82%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Low

Low-risk findings worth noting

SKILL.md
Quality
Evals
Security

Quality

Content

65%

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

The body is highly actionable with specific commands, paths, and diagnostic tells, and is well organized, but it is long and monolithic with no progressive disclosure to bundle files. Some diagnostic prose could be tightened and workflows framed as explicit checkpointed sequences.

Suggestions

Move the detailed Common Diagnostics cases and the Pointers file list into reference bundle files (e.g. references/diagnostics.md, references/file-map.md), keeping only the top tells and entrypoints inline, to improve progressive_disclosure and conciseness.

Tighten the Common Diagnostics prose (e.g. the input-available / persistIncomingTail explanation) to the minimum signal needed, trimming redundant restatement.

Frame the verification and diagnostic flows as short numbered checkpointed sequences (e.g. 1. typecheck+test touched package -> 2. rebuild daemon if touched -> 3. agent typecheck -> 4. smoke:sessions:live) to make workflow checkpoints explicit.

DimensionReasoningScore

Conciseness

The body mostly earns its tokens with project-specific architecture, boundaries, and verification commands and does not explain concepts Claude already knows, but the Common Diagnostics prose is long and could be tightened without losing the signal.

2 / 3

Actionability

Provides concrete, executable guidance: copy-paste pnpm verification commands, a read-only sqlite3 query, and specific source paths plus pinpoint fixes (e.g. "add the offender to IGNORED_SCAN_DIRS in @grida/daemon's workspaces/scan.ts").

3 / 3

Workflow Clarity

Sequenced procedures exist (verification pair with the build-order checkpoint that @grida/daemon must rebuild before @grida/agent typechecks, and a "first thing to check" diagnostic flow), but they are presented as scattered reference prose rather than explicit checkpointed validate-fix-retry loops.

2 / 3

Progressive Disclosure

Sections are well organized (Shape, Boundaries, Ownership, Diagnostics, Live state, Security, Verification, Pointers) and references are one level deep, but the ~290-line body is monolithic with no bundle files to offload detailed diagnostics or the full pointers list.

2 / 3

Total

9

/

12

Passed

Description

100%

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 specific, trigger-rich, and complete, with explicit use-for guidance and a clean exclusion to a sibling skill. It is a strong, low-conflict description.

DimensionReasoningScore

Specificity

Lists many concrete capabilities across both packages (DaemonServer, loopback HTTP perimeter, files/workspaces, secrets store, daemon discovery, sessions, providers/BYOK, runtime/tool execution, skills discovery, prompts, tiers, sandbox hosts) rather than vague language.

3 / 3

Completeness

Answers both what (the two packages and their responsibilities) and when with an explicit "Use for ..." trigger list plus an explicit exclusion clause ("For pure Electron window, preload, menu, deep-link, or CDP work, use desktop").

3 / 3

Trigger Term Quality

Uses path- and surface-based triggers a developer on this codebase would naturally say ("packages/grida-daemon/**", "packages/grida-ai-agent/**", "desktop sidecar protocol", "agent chat transport", "agent state or streams").

3 / 3

Distinctiveness Conflict Risk

Has a clear niche (the daemon + agent tenant packages) and an explicit hand-off to the sibling "desktop" skill, making overlap with other skills unlikely.

3 / 3

Total

12

/

12

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.

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

relative_links

Relative link issues: 4 suspicious

Warning

Total

15

/

16

Passed

Repository
gridaco/grida
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.