ml-impl

A workflow for turning an approved specification into code through planning, implementation, and review. It uses the specification, related wireframes, existing project patterns, and shared utilities as context.

In plain words
What is it for?
Use it to implement documented screens or components, inspect similar existing code, follow project styling, and review the completed implementation.
Why use it?
It provides a structured path from an agreed design to working code and encourages checking the result against the specification and the rest of the codebase.

Command

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/jpavley/meta-loop-ios/ml-impl
Clone the repo
git clone --depth 1 https://github.com/jpavley/meta-loop-ios
Per session 25 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,057 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 $0.00025 $0.02057
Opus 5 $0.00013 $0.01028
Sonnet 5 $0.00005 $0.00411
Haiku 4.5 $0.00003 $0.00206

Measured yesterday against content hash d349707e283b, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

ml-impl 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.

commands/ml-impl.md · 344 lines

How it starts

The opening of the file, as written. The whole thing — 344 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Meta-Loop: Implementation Loop

Complete the implementation loop: plan the implementation, write code, and review for correctness.

Modes: Plan Mode → Agent Mode → Ask Mode

Arguments

  • $ARGUMENTS: Path to the specification to implement
    • Examples: specs/views/my-feature/spec.md

Prerequisites

This skill implements Steps 3.1-3.4 of the Meta-Loop process defined in skills/meta-loop-methodology/SKILL.md.

The specification should have already passed /ml-spec review before invoking this skill.


Phase 1: Plan (Plan Mode)

Enter plan mode and create a plan for implementing the specification.

1.1 Load Context

  1. Load the specification from $ARGUMENTS
  2. Load associated wireframes from the same directory
  3. Load notes.md from the same directory (if it exists)
    • This file contains decisions from wireframe and spec phases
    • Clarifying questions and their answers
    • Key decisions and technical notes
  4. Read CLAUDE.md for codebase patterns and conventions
  5. Load referenced view specs (follow links in the spec)
  6. Explore relevant existing code:
    • Similar views for patterns
    • Shared utilities and helpers
    • Theme and styling configuration

1.2 Create the Plan

The plan should outline:

Files to Create:

  • Full paths for new files
  • Purpose of each file

Files to Modify:

  • Full paths for existing files
  • Summary of changes to each

Implementation Order:

  1. [First step - usually models/types]
  2. [Second step - usually core logic]
  3. [Third step - usually UI components]
  4. [Fourth step - usually integration]
  5. [Fifth step - usually tests]

Testing Strategy:

  • Unit tests for logic
  • Preview support for UI components
  • Integration points to verify

Build Verification:

  • How to verify the build succeeds
  • What to test manually

1.3 Present Plan to Developer

Present the plan clearly and wait for developer feedback.

## Implementation Plan

### Specification
`$ARGUMENTS`

### Files to Create
| Path | Purpose |
|------|---------|
| `path/to/NewFile.swift` | [Purpose] |
| ... | ... |

### Files to Modify
| Path | Changes |
|------|---------|
| `path/to/ExistingFile.swift` | [Brief summary] |
| ... | ... |

### Implementation Order
1. **[Step]:** [Details]
2. **[Step]:** [Details]
...

### Testing Strategy
- [ ] [Test approach 1]
- [ ] [Test approach 2]

### Questions for Developer
- [Any clarifications needed before implementing]

Read the full file on GitHub · 344 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. yesterday First seen · 344 lines · 25 tokens per session scan A d349707e283b

Subscribe to this mod's changes

ml-impl is a command published in the GitHub repository jpavley/meta-loop-ios (3 stars, last pushed 6mo ago), licensed MIT. It adds 25 tokens to every session and 2,057 once invoked, about $0.0001 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 2026-08-31.