agentmods.dev

finish-him

finish-him is a command for Claude Code from henchmarketing-rgb/sub-zero-skill. It costs 32 tokens per session (10,634 once invoked), scanned A, original, MIT.

A two-stage project completion command. It first audits the codebase and creates a plan, then can execute that plan step by step with checks against the real project or deployed result.

In plain words
What is it for?
Use it to inspect incomplete work, produce an HTML and JSON plan, carry out planned changes, verify each step, and retry failed steps.
Why use it?
It turns an unfinished project into a tracked path toward a defined outcome and catches claims of success that are not supported by actual verification.

Command for Claude Code

Written for Claude Code: ${CLAUDE_PLUGIN_ROOT} variable. Also seen: mentions CLAUDE.md; mentions subagents; positional $N argument.

Runs only inside its plugin — its command needs a path that Claude Code sets for a plugin’s own hooks and for nothing else. Install the plugin, not this.

Part of the sub-zero plugin — 1 command, 2 agents shipped together

not rated 5repo 3mo ago A scan 32 tokens original MIT

Good fit Use it to inspect incomplete work, produce an HTML and JSON plan, carry out planned changes, verify each step, and retry failed steps.

Compare 6 commands from other repositories ↓
View source ↗ henchmarketing-rgb/sub-zero-skill
Install

Getting it into your agent

This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.

Claude Code
/plugin marketplace add henchmarketing-rgb/sub-zero-skill
Claude Code
/plugin install sub-zero

Made for: Claude Code.

Or install sub-zero, the plugin that ships this one along with the rest of its 1 command, 2 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 finish-him

README.md
[![agentmods](https://agentmods.dev/badge/commands/henchmarketing-rgb/sub-zero-skill/finish-him/github.svg)](https://agentmods.dev/commands/henchmarketing-rgb/sub-zero-skill/finish-him)
Your own site
<a href="https://agentmods.dev/commands/henchmarketing-rgb/sub-zero-skill/finish-him"><img src="https://agentmods.dev/badge/commands/henchmarketing-rgb/sub-zero-skill/finish-him/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for finish-him

Your own site · 80×15
<a href="https://agentmods.dev/commands/henchmarketing-rgb/sub-zero-skill/finish-him"><img src="https://agentmods.dev/badge/commands/henchmarketing-rgb/sub-zero-skill/finish-him.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 32 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 10,634 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. A grade says what 26 rules found in the file — not that it is safe.
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 $0.00032 $0.10634
Opus 5 $0.00016 $0.05317
Sonnet 5 $0.00006 $0.02127
Haiku 4.5 $0.00003 $0.01063

Measured 10d ago against content hash c169c6dcd581, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-10, from the pricing page.

Security

Grade A, and why

finish-him scanned grade A with 1 finding 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 10d 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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

echo "$url: $(curl -sI -m 5 "$url" | head -1)"
commands/finish-him.md · 694 lines

How it starts

The opening of the file, as written. The whole thing — 694 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/finish-him

You are Sub-Zero. Two modes, one slash command.

  • Mode A (audit) — default invocation. Read the project deep, find what's done / broken / incomplete / stale / missing, write a multi-step plan from current state to the frozen win condition, hand the operator an HTML report with a button to launch Mode B. NO code changes, NO commits, NO deploys.
  • Mode B (execute) — triggered by --execute=<plan-path>. Read the plan JSON, execute each step on operator ship it, dispatch verifier per step, debug-loop on FAIL, finalize the HTML with a YOU WIN!!! or YOU LOSE!!! closer.

State to track across a run

Hold these in memory:

  • mode — audit or execute
  • run_timestamp — YYYY-MM-DD-HHmm, set in §1
  • project_slug, project_root
  • win_condition — frozen string for this run (resolved in §0 from the command arg or .finish-him/win.md)
  • plan_html_path, plan_json_path — <project_root>/docs/finish-him/<ts>-<slug>-plan.{html,json}
  • verifier_history, commit_history, debug_history, step_history
  • dispatch_count — increment per sub-agent call (§Circuit breaker)

0. Mode detection and win-condition resolution

Parse args in this precedence order. --project=<slug> is a flag (overrides slug detection in either mode); pull it out first, then:

  1. --execute=<path> present → mode = "execute", plan_json_path = <path>. Jump to §B1. (The frozen win condition is read from the plan JSON, not re-resolved.)
  2. Else a non-flag positional argument is present → mode = "audit", win_condition = <that argument, verbatim>. Continue to §A1.
  3. Else .finish-him/win.md exists in the project root → mode = "audit", win_condition = <trimmed contents of that file>. Continue to §A1.
  4. Else → print No win condition. Pass one as an argument (/finish-him "your win condition") or create .finish-him/win.md in the project root, then re-run. Exit.

The win condition is plain text. Example: /finish-him "homepage live at production URL and links to the docs".

Read the full file on GitHub · 694 lines

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. 10d ago First seen · 694 lines · 32 tokens per session scan A c169c6dcd581

Subscribe to this mod's changes

finish-him is a command published in the GitHub repository henchmarketing-rgb/sub-zero-skill (5 stars, last pushed 3mo ago), licensed MIT. It adds 32 tokens to every session and 10,634 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Claude CodeDebuggingPlanningai-agentsclaudedeveloper-toolsverification
Related

Other commands, from other repositories

ui-aqa-flow-test-report-analysis

Phase 7 Test Report Analysis of ui-aqa-flow.

griddynamics/rosetta · 19 tokens

api-aqa-flow-execution-and-report-analysis

Phase 6 Execution & Report Analysis of api-aqa-flow (USER INTERACTION REQUIRED).

griddynamics/rosetta · 28 tokens

fire-resurrect

Autonomous Resurrection Mode — reverse-engineer intent from messy code, then rebuild clean from scratch.

ThierryN/fire-flow · 18 tokens

fire-reflect

After any failure (debug resolution, test failure, approach rotation, stalled loop), capture what was tried, why it failed, and what actually worked as a persistent reflection. Future sessions search these before investigating.

ThierryN/fire-flow · 3 tokens

fire-verify-uat

Conversational User Acceptance Testing with automatic parallel diagnosis on failures.

ThierryN/fire-flow · 11 tokens

sruja-bugfix-triage

Systematic workflow for triaging and fixing bugs with architecture grounding.

sruja-ai/sruja · 0 tokens