Avenir-Web

Trust Score 66/100

Run and improve Avenir-Web autonomous web tasks: choose run mode, validate and refine instructions, execute single or batch runs, and analyze outputs to propose

triggers:run a task on a websiteatomic actionread pagebatch experimentheadlessdemo mode

GitHub SKILL.md

What it does

Avenir-Web provides a CLI-driven runtime for executing and iterating on autonomous web tasks. It supports single tasks, atomic browser actions, read-page queries (screenshot + model question), and batch experiments. The skill helps validate instruction quality, select an appropriate run mode (headless/headed/demo), execute runs, and synthesize results with a recommended next action.

When to use it

Use this skill when you need to: run a deterministic web automation task, perform one precise browser action, inspect a page via screenshot and ask a model a question, or run and summarize large batch experiments with reproducible configuration. Ideal for debugging, benchmarking, and improving task success rates.

What's included

Scripts: canonical CLI entrypoints (example.py, scripts/atomic_action.py, scripts/read_page.py, src/run_agent.py for batches) described in the SKILL.md body.
References: usage templates, batch config checklist, and output/report contract included in the skill file.
Instructions: guidance on instruction design (objective, constraints, completion condition), mode selection rules, and reporting fields (task_id, status, evidence, cause, next step).

Compatible agents

Compatible with agents capable of running Python CLIs and browser automation (Playwright-based agents, Codex/GitHub Copilot-style runners, or any agent with shell and browser control).

Audit Summary

Avenir-Web is a web automation skill that runs single/batch browser tasks via Playwright with an OpenRouter-powered agent. Both bundled scripts (atomic_action.py, read_page.py) failed immediately due to missing 'avenir_web' module — they depend on the full repo being installed. SKILL.md is well-structured with clear workflows and mode guidance, but scripts cannot run standalone outside the repo environment.

Watch Out

Requires full avenir_web repo installed (pip install -e src) — scripts are not self-contained
Needs OPENROUTER_API_KEY and Playwright Chromium installed
Scripts modify sys.path to find local src/ directory

Missing Dependencies

avenir_webplaywrighttoml

Notes

Security: No hardcoded credentials, no destructive commands, no exfiltration. Scripts reference an external API key (OPENROUTER_API_KEY) but instruct not to hardcode it — good practice. Deductions for shell injection risk from unquoted URL args passed to subprocess-like agent calls (-12), and network calls to arbitrary user-provided URLs via Playwright agent (-10), and sys.path manipulation (-6). Code quality: Scripts are clean async Python with argparse, proper error handling via try/finally, and structured JSON output. However they crash on import outside the repo — not idempotent or portable. SKILL.md has clear workflows but scripts can't be run independently. Architecture: Frontmatter is minimal (name + description only, missing metadata). Scripts live in scripts/ which is good, but they deeply depend on the repo's src/ package. Output contracts are defined but only work within the full repo context. Usefulness: Web automation via Playwright is a real need, but the skill requires significant setup (repo install, Playwright, API key) and targets a specific academic project rather than a general-purpose tool. Niche audience of researchers/developers working with the Avenir-Web framework.

Information

Repository: avenir-web
Stars: 29

Trust Score

Overall66

Security72

Code Quality62

Architecture55

Usefulness58

Related Skills

ds-fix — data-science mid-analysis fixer

Orchestrates diagnosis and targeted fixes mid-analysis: diagnose root cause, apply fixes with output-first verification, and update project learnings.

Readwise Reader Document Management

Manage Readwise Reader documents: list, save, search, move, tag, highlight, export and bulk-edit via official and custom CLIs.

ezBookkeeping API Tools

Command-line API tools for ezBookkeeping: record and query transactions, retrieve accounts/categories/tags, and fetch exchange rates for self-hosted personal fi

Tmux Bridge

Give an agent controlled access to a local tmux terminal bridge so it can run shell commands, capture output, and manage sessions on the user's machine.

SourceSage CLI

Generate concise, LLM-friendly repository summaries (Markdown) using the SourceSage CLI — supports lite mode, language switching, and targeted repo analysis.

Nit (Nitter CLI)

Terminal client to browse tweets, view profiles, and search posts via Nitter instances without opening a browser.

hn — Hacker News CLI

Terminal-based Hacker News client with a CLI and optional interactive TUI for browsing top/new/best/ask/show/jobs stories, viewing threads with nested comments,

Lokalise Migration — Deep Dive

Step-by-step migration guide and tooling for moving translations from Crowdin/Phrase/POEditor into Lokalise, including transforms, uploads and validation.

Back to Skills