Goal Packet Writer

    1

    Turns a vague goal into a packet: allowed files, runnable checks, a budget and stop rules, then validates its form.

    $50

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Goal Packet Writer

    Example session with this skill installed

    The greet function in hello.py should trim spaces around the name before building the greeting, so greet(' Ann ') returns 'Hello, Ann!'. Write a brief for another agent to do this work. Do not do the work yourself.

    • Read your context and instructions
    • Compiled the goal packet writer

    Summary

    The agent wrote a packet with the seven sections, ran the form check, and showed it to you. It did not start the work. Output below is from running the checker on the packet.

    The packet

    
    # Packet: greeting
    
    ## Goal
    Add a greet(name) function to hello.py that returns "Hello, NAME!".
    
    ## Context files
    - README.md
    
    ## Allowed files
    - hello.py
    - test_hello.py
    
    ## Out of scope
    - Any file not listed above
    - Packaging or CI
    
    ## Acceptance checks
    - `python3 -m unittest` passes
    - `git diff --stat` lists only the two allowed files
    
    ## Budget
    30 minutes and 40 turns.
    
    ## Stop and ask
    - A test needs a file that is not allowed
    - The same error twice
    

    Form check

    $ python3 scripts/check_packet.py PACKET.md
    PASS packet form is complete
    

    A weak packet (goal "Improve the code.", allowed files **, a check with no command, budget "Soon.") gives:

    WARN Goal: vague word: Improve
    FAIL Allowed files: too broad or a pattern: **
    FAIL Out of scope: add the bullet "Any file not listed above"
    FAIL Acceptance checks: no check names a command in backticks
    FAIL Budget: give a number with a unit (seconds, minutes, hours, turns, tokens or money)
    

    Next step

    Read the packet. If it is right, say yes and the work can start. When it is done, each acceptance check is run and its real output reported.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    A vague goal gets a vague result. This skill makes the agent write a packet before any work starts.

    The packet has seven sections: Goal (one or two sentences), Context files, Allowed files (no wildcards), Out of scope, Acceptance checks (each one a command), Budget (a number with a unit) and Stop and ask.

    A script checks the form: the seven sections present once and in order, a short goal, no vague words (warning), no wildcard or absolute file paths, at least one check with a command in backticks, and a positive number with a unit in the budget. The agent shows you the packet and waits for your yes. When the work is done, it runs each check and reports the real output.

    Standard library only. Written for Python 3.10 or later, tested on 3.13 only. 15 tests.

    What it does not do It checks the form, not whether the goal is a good idea or the checks prove it. A passing packet can still be a bad plan. It does not run the work.

    Before you use this These skills are new and have not been tested widely. They come from a project that works and that I use. You are responsible for what you run and for reviewing what your agent does with them, so check results before you rely on them. The author is not liable for loss or damage, in line with the MIT licence.

    Requires python3 to run the internal check_packet.py script. Cannot verify if the goal is logically sound, only if the packet structure is valid.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed today

    Requires

    Python 3
    Git

    Permissions

    Terminal / Shell
    Read Files
    Write Files

    No network access. It writes a PACKET.md file where you say, or in a temporary folder if you name none. The validator only reads that file.

    Listedtoday

    What's inside

    Frequently Asked Questions

    Seven sections: Goal, Context files, Allowed files, Out of scope, Acceptance checks, Budget, and Stop and ask. The agent fills each one from your goal and from facts it reads in your project.

    The form only: the seven sections present once and in order, a goal of one or two sentences, no wildcard or absolute file paths, the "Any file not listed above" bullet, at least one check with a command in backticks, and a positive number with a unit in the budget.

    No. It writes the packet and checks its form, then waits for your yes. When the work is done by you or an agent, the acceptance checks are run and their real output is reported.

    No. The validator cannot tell whether the goal is a good idea or whether the checks prove it. Read the packet before you say yes.

    The validator is a Python script that runs anywhere, and the instructions are written for agents that read SKILL.md. I have tested it with Claude Code only.

    The SKILL.md instructions, the validator script (check_packet.py), an examples file and 15 unit tests. Standard library only.

    New in Agents & Orchestration