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.
npx agentmods add commands/dagster-io/dagster/plan-implementgit clone --depth 1 https://github.com/dagster-io/dagsterWhat 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 | $0.00016 | $0.02028 |
| Opus 5 | $0.00008 | $0.01014 |
| Sonnet 5 | $0.00003 | $0.00406 |
| Haiku 4.5 | $0.00002 | $0.00203 |
Grade C, and why
plan-implement scanned grade C 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 2d 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.
Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
git rm -rf .worker-impl/ How it starts
The opening of the file, as written. The whole thing — 288 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/erk:plan-implement
Implement a plan - either from a GitHub issue, a markdown file, an existing .impl/ folder, or by saving the current plan first.
This is the primary implementation workflow - it orchestrates:
- Setting up the
.impl/folder (from issue, file, existing folder, or fresh plan) - Executing the implementation
- Running CI and submitting the PR
Prerequisites
- Must be in a git repository managed by erk
- GitHub CLI (
gh) must be authenticated - One of:
- An issue number, URL, or file path argument
- An existing
.impl/folder - A plan in
~/.claude/plans/(from plan mode)
Usage
/erk:plan-implement # Use .impl/ or save current plan
/erk:plan-implement 2521 # Fetch and implement issue #2521
/erk:plan-implement https://github.com/owner/repo/issues/2521 # URL form
/erk:plan-implement ./my-plan.md # Implement from local markdown file
Agent Instructions
Step 0: Parse Arguments
Extract optional argument from $ARGUMENTS:
- If numeric (e.g.,
2521): Store asISSUE_ARG - If GitHub URL (e.g.,
https://github.com/owner/repo/issues/2521): Extract number from path, store asISSUE_ARG - If path to file (anything else non-empty): Store as
FILE_ARG - If empty: Proceed to check
.impl/folder
Store either ISSUE_ARG (issue number) or FILE_ARG (file path), or neither if empty.
Step 1: Determine Implementation Source
Follow this priority order:
1a. If ISSUE_ARG is provided
Set up from the specified issue:
erk exec setup-impl-from-issue <ISSUE_ARG>
This command:
- Creates a feature branch from current branch (stacked) or trunk
- Checks out the new branch in the current worktree
- Creates
.impl/folder with the plan content - Saves issue reference for PR linking
If this fails, display the error and stop.
Then run impl-init:
erk exec impl-init --json
1a-file. If FILE_ARG is provided
Set up from the specified markdown file:
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.
- 2d ago First seen · 288 lines · 16 tokens per session scan C cdc094cfbeb6
plan-implement is a command published in the GitHub repository dagster-io/dagster (16,073 stars, last pushed 4d ago), licensed Apache-2.0. It adds 16 tokens to every session and 2,028 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it C with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
OPSX: Continue
Continue working on a change - create the next artifact (Experimental).
OPSX: Fast Forward
Create a change and generate all artifacts needed for implementation in one go.
add-task-module
Scaffold a new task module (classification/regression/clustering/...).
work-on-approved-issue
Pick an Approved GitHub issue and resolve it end-to-end.
run-engine-tests
Fast local test loop for the engine package.
cron
Schedule skills to run automatically at specific times or based on events.