A command for executing one approved phase from a documented implementation plan. It builds only that phase, runs its checks, commits the result to a feature branch, and stops before opening a pull request.

Use it to implement a single plan phase, run phase tests plus pytest and ruff, update the progress tracker, and prepare a verified commit.

It keeps implementation limited to an agreed scope and requires verification before handing the work to the next delivery step.

implement — Command by littlebearapps

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 commands/littlebearapps/untether/implement

Clone the repo

git clone --depth 1 https://github.com/littlebearapps/untether

Made for: Claude Code.

Per session 78 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 1,593 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.

Model	Per session	Once invoked
Fable 5	$0.00078	$0.01593
Opus 5	$0.00039	$0.00796
Sonnet 5	$0.00016	$0.00319
Haiku 4.5	$0.00008	$0.00159

Measured yesterday against content hash 83e3ddb209e4, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

implement 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 yesterday.

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.

.claude/commands/implement.md · 138 lines

How it starts

The opening of the file, as written. The whole thing — 138 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are handling /implement. /implement executes exactly one approved phase of a docs/plans/<slug>/ pack to a verified hand-off, then STOPS before the PR (/pr-dev opens it). It is the build half of delivery.

User input: $ARGUMENTS

Untether adaptations (read first)

Load .claude/rules/workflow-commands.md (routing + cross-cutting rules). For /implement:

Authority: code + tests in the approved phase scope only, plus the one progress-tracker.md 🟡 mark. Never opens the PR. Never widens scope (→ STOP + ask). Never takes defect work (→ /fix). Never runs paid work (N/A).
Release-guard. Commit to feature/<slug>-phase-N (branch off dev). Never master/tag/release. Never --no-verify.
Dev/staging. Verify on untether-dev; never restart staging. Never systemctl restart from inside this session.
Refuse multi-phase. One phase per invocation, always.

Reuse (compose, don't re-describe)

Via the Skill tool: superpowers:executing-plans (phase discipline) + superpowers:test-driven-development (red→green→refactor) + superpowers:verification-before-completion (evidence before "done"). If the phase touches a runner/schema/telegram surface, also load the matching .claude/rules/*.md and the relevant .claude/skills/*.

Sub-commands

Form	Action
`/implement <slug> <phase-N>`	Execute that phase (TDD), commit, stop before PR
`/implement <slug>`	Read-only: print the next-up phase + its pre-flight status
`/implement <slug> <phase-N> --dry-run`	Write the failing tests + stage locally; no branch push
`/implement --help`	Usage, then stop

Flow

I-1. Gate check

Resolve docs/plans/<slug>/ and the named phase file. If the pack or phase is missing → STOP (route to /plan).
Confirm the phase is approved and its dependencies (earlier phases) are ✅ in progress-tracker.md. If a dependency is unmet → STOP and say which.
If more than one phase is named → refuse (one phase only).

Read the full file on GitHub · 138 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.

yesterday First seen · 138 lines · 78 tokens per session scan A 83e3ddb209e4

Subscribe to this mod's changes

implement is a command published in the GitHub repository littlebearapps/untether (66 stars, last pushed 2d ago), licensed MIT. It adds 78 tokens to every session and 1,593 once invoked, about $0.0004 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.

Code review Git and workflow Testing ai-agents claude-code cli-tools codex developer-tools opencode

implement

Getting it into your agent

What it costs to keep this loaded

Grade A, and why

Nothing flagged

How it starts

Untether adaptations (read first)

Reuse (compose, don't re-describe)

Sub-commands

Flow

I-1. Gate check

What this file has done since we first saw it

Other commands, from other repositories

psm

release

release

self-improve

research

wtfp:discuss-section