Tessl • The management layer for AI agent skills

Blog Docs Log inGet started

SkillPublic

agent-browser

Automates browser and Electron app interactions for user-flow validation.

Quality

70%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./.agents/skills/agent-browser/SKILL.md

Quality

Content

73%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 an exceptionally actionable, well-sequenced CLI reference with strong validation patterns (diff, re-snapshot, troubleshooting), and it avoids explaining concepts Claude already knows. Its main weaknesses are inlining everything into one ~765-line file with zero progressive disclosure, and duplicated coverage of connection/debugging options across sections.

Suggestions

Split the command catalog (Commands, Global options, Environment variables, Configuration file) into a references/ file such as commands.md, keeping only the quick start, core workflow, and ref lifecycle in SKILL.md.

Move the Electron automation guide (~200 lines, including platform-specific launch commands and troubleshooting) into a separate references/electron.md referenced one level deep.

Deduplicate the --cdp/connect/--auto-connect documentation, which currently appears under Connect to existing Chrome, Connecting to Electron Apps, and Debugging.

Dimension	Reasoning	Score
Conciseness	The command listings are dense and free of concept explanations Claude already knows, but there is notable redundancy: --cdp/connect/--auto-connect are documented three times (Connect to existing Chrome, Connecting to Electron Apps, Debugging) and re-snapshot guidance is repeated across several sections, so the body could be meaningfully tightened.	3 / 5
Actionability	Every command is copy-paste ready (e.g., `agent-browser fill @e1 "user@example.com"`, the eval --stdin heredoc pattern) and worked examples cover the common cases — form submission, authentication with saved state, annotated screenshots — plus concrete troubleshooting fixes like `lsof -i :9222` and `sleep 3`.	5 / 5
Workflow Clarity	The core loop (open → snapshot -i → interact by ref → re-snapshot) is explicitly sequenced, the ref-lifecycle section marks invalidation checkpoints, and the diff workflow provides an explicit validate-after-action feedback loop (snapshot → action → diff snapshot) with error-recovery guidance in the troubleshooting sections.	5 / 5
Progressive Disclosure	Although section headers are clear, the entire ~765-line reference — full command catalog, Electron automation guide, config/env-var reference — is inlined in a single SKILL.md with no bundle files or pointers, matching anchor 2's 'content that clearly belongs in separate files is inlined' rather than anchor 3, which expects at least some references to be present.	2 / 5
	Total	15 / 20 Passed

Description

53%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 concise and names a distinct niche with a clear purpose, but it stops at one generic action verb and provides no 'when to use' guidance or comprehensive trigger terms. Adding an explicit 'Use when...' clause with concrete triggers would address the main gaps.

Suggestions

Add an explicit trigger clause, e.g., "Use when the user asks to drive a browser, test a web UI or user flow, take page screenshots, or automate an Electron desktop app."

Enumerate concrete capabilities to raise specificity, e.g., "Navigate pages, click/fill elements via accessibility snapshots, take annotated screenshots, record video, and diff page states."

Include common synonyms users would say — web automation, UI testing, scraping, Chrome, Playwright-style flows — to broaden natural keyword coverage.

Dimension	Reasoning	Score
Specificity	The phrase "Automates browser and Electron app interactions" names the domains and one generic action, and "for user-flow validation" gives a purpose, but there are no distinct concrete actions (e.g., click, fill, screenshot) as required for anchor 4.	3 / 5
Completeness	There is a clear 'what' ("Automates browser and Electron app interactions") but no 'Use when...' clause or equivalent explicit trigger guidance; 'for user-flow validation' only weakly implies 'when', so completeness is capped at 3 per the judging guidelines.	3 / 5
Trigger Term Quality	"browser", "Electron", and "user-flow validation" are natural keywords, but common variations users would say — screenshot, scraping, UI testing, Chrome, form filling — are missing, matching anchor 3's 'relevant keywords but missing common variations'.	3 / 5
Distinctiveness Conflict Risk	The browser/Electron automation niche is mostly distinct with only minor overlap risk against closely related web-scraping or testing skills; it is not a 5 because no explicit trigger phrases delineate it from those neighbors.	4 / 5
	Total	13 / 20 Passed

Validation

93%

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 — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (765 lines); consider splitting into references/ and linking	Warning

	Total	15 / 16 Passed

Repository: jomi-se/bookhand
Path: .agents/skills/agent-browser/SKILL.md
Revision: b7ee13d

Reviewed: 4 days ago

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.

agent-browser

Table of Contents

Is this your skill?