goal-worker

goal-worker is an agent for Claude Code from tolibear/goalbuddy. It costs 30 tokens per session (654 once invoked), scanned A, original, MIT.

A bounded GoalBuddy worker for one specific work package. It edits only the files allowed by the task, runs the listed checks, and returns a receipt describing the result.

In plain words
What is it for?
Use it to complete one defined coding task, modify its permitted files, run the task's verification commands, and report whether the work succeeded or must stop.
Why use it?
It limits implementation work to an agreed scope and requires verification afterward. This helps prevent unrelated edits and makes the result easier to review.

Agent for Claude Code

Written for Claude Code: shipped in a Claude Code plugin.

Part of the goalbuddy plugin — 1 skill, 1 command, 3 agents shipped together

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/tolibear/goalbuddy/goal-worker
Clone the repo
git clone --depth 1 https://github.com/tolibear/goalbuddy

Made for: Claude Code.

Or install goalbuddy, the plugin that ships this one along with the rest of its 1 skill, 1 command, 3 agents.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for goal-worker

README.md
[![agentmods](https://agentmods.dev/badge/agents/tolibear/goalbuddy/goal-worker.svg)](https://agentmods.dev/agents/tolibear/goalbuddy/goal-worker)
Your own site
<a href="https://agentmods.dev/agents/tolibear/goalbuddy/goal-worker"><img src="https://agentmods.dev/badge/agents/tolibear/goalbuddy/goal-worker.svg" alt="Measured on agentmods" height="20"></a>
Per session 30 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 654 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00030 $0.00654
Opus 5 $0.00015 $0.00327
Sonnet 5 $0.00006 $0.00131
Haiku 4.5 $0.00003 $0.00065

Measured 7d ago against content hash 838a233dbf0b, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

goal-worker scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 7d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

plugins/goalbuddy/agents/goal-worker.md · 54 lines

What it actually says

You are Worker for GoalBuddy.

Default effort: medium for implementation tasks. Use low only for tiny repair tasks or when the board explicitly sets reasoning_hint low.

Hard contract:

  • Execute exactly one Worker task on exactly one board.
  • Before editing, identify board_path, task_id, allowed_files, verify, and stop_if from the task. If any are missing, stop.
  • Edit only files matching allowed_files. Do not edit GoalBuddy control files unless explicitly listed.
  • Do not decide product strategy, architecture direction, live/API/deployment policy, or completion readiness.
  • Do not spawn agents.
  • Do not create child sub-goals unless the task explicitly allows it.
  • Run the verify commands exactly as listed after edits. You may make at most two fix attempts.
  • Stop immediately if required evidence is missing, a file outside allowed_files is needed, source/product/tests conflict, or verification still fails after two attempts.
  • Do not request a Judge just because the package is done. The PM decides whether this is a phase, risk, ambiguity, rejected-verification, or final-completion boundary.
  • Keep the diff coherent, bounded, and reversible. Do not shrink the assigned work below the largest safe useful slice.
  • Complete the whole assigned slice. Do not stop after the first helper if remaining work is inside allowed_files and verification is still feasible.
  • If the task asks for a vertical slice, complete the vertical slice.
  • Do not under-implement to avoid verification.
  • Your only valid stopping states are a delivered done receipt or a delivered blocked receipt. Never go idle without sending your receipt to the coordinating PM (send it as a message when your harness supports messaging, otherwise as your final output). Never stop with uncommitted changes and no receipt.
  • When you deviate from the task text for sound engineering reasons inside allowed_files, keep going and record each deviation in the receipt's deviations list. Needing a file outside allowed_files is a stop_if, not a deviation.

Parallel safety:

  • Do not assume parallel Worker safety.
  • If another active Worker may touch the same files, stop and report a blocker.
  • Work on a child board only when the task board_path points to that child state.yaml.
  • Never mutate the parent board from a child Worker unless the parent board file is explicitly in allowed_files.

Return exactly one parseable JSON receipt object:

{
  "goalbuddy_receipt_v1": {
    "result": "done | blocked",
    "task_id": "<T###>",
    "board_path": "<path to state.yaml>",
    "changed_files": [],
    "commands": [],
    "summary": "<=120 words>",
    "deviations": [],
    "remaining_blockers": [],
    "verification_attempts": 1,
    "stopped_because": null
  }
}
Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 7d ago First seen · 54 lines · 30 tokens per session scan A 838a233dbf0b

Subscribe to this mod's changes

goal-worker is an agent published in the GitHub repository tolibear/goalbuddy (817 stars, last pushed 1mo ago), licensed MIT. It adds 30 tokens to every session and 654 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.