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/drobins25/craft/craft-cycle-startgit clone --depth 1 https://github.com/drobins25/craftWhat 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 | $0.00012 | $0.02638 |
| Opus 5 | $0.00006 | $0.01319 |
| Sonnet 5 | $0.00002 | $0.00528 |
| Haiku 4.5 | $0.00001 | $0.00264 |
Grade A, and why
cycle-start scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 345 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Cycle Start
Activate a cycle and begin implementation.
Flow
Step 1: Select Cycle
If cycle name provided:
Verify it exists in .craft/cycles/[name]/
If no cycle specified:
Scan for cycles that can be activated (planning or ready status only):
# Discover activatable cycles
Glob ".craft/cycles/*/cycle.yaml" → cycle_yaml_files
activatable_cycles = []
for each cycle_yaml in cycle_yaml_files:
Read cycle_yaml → parse status, title
if status == "planning" or status == "ready":
cycle_name = directory name from path
cycle_title = title value
Glob ".craft/cycles/$cycle_name/stories/*.md" → count → story_count
activatable_cycles.append(cycle_name, cycle_title, story_count, status)
Sort and recommend:
- Extract numeric prefix from cycle directory names (e.g.,
1-auth→1,2-dashboard→2) - Sort cycles by numeric prefix (lowest first)
- The recommended cycle is the lowest-numbered cycle that isn't
complete - If cycles lack numeric prefixes, present them in filesystem order without recommendation
Recommendation logic:
# For each cycle, extract prefix: cycle_name | sed 's/^\([0-9]*\)-.*/\1/'
# Sort by prefix number
# First in sorted list = recommended
"Which cycle do you want to start?"
Use AskUserQuestion (options sorted by numeric prefix, recommended marked):
question: "Which cycle do you want to start?"
header: "Cycle"
options:
- label: "[1-auth] — Auth Flow (Recommended)"
description: "3 stories, ready — next in sequence"
- label: "[2-dashboard] — Dashboard Foundation"
description: "6 stories, ready"
- label: "[3-api] — API Layer"
description: "4 stories, [planning] — needs detailing first"
Note: For cycles with status: planning, include [planning] in the description to signal they may need detailing. Step 2b already handles the case where stories aren't ready.
Note: Show "Already active: [cycle]" in the question text if applicable.
Note: If cycles don't have numeric prefixes (legacy projects), omit the "(Recommended)" marker and present cycles in the order found.
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 · 345 lines · 12 tokens per session scan A c630bd138ae9
cycle-start is a command published in the GitHub repository drobins25/craft (52 stars, last pushed 2d ago), licensed MIT. It adds 12 tokens to every session and 2,638 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 2026-08-30.
Other commands, from other repositories
implement
Execute phased implementation with validation gates.
README
Git workflow and quality assurance commands for the claude-skills repository.
output
Generate output artifacts from active wiki content — summaries, reports, study guides, slide outlines, timelines, glossaries, comparisons. Outputs are filed back into the wiki.
review
Code review with security and @MX tag compliance check.
analyst
Use when performing local analyst review before pushing PR changes. Assesses code quality, impact analysis, and maintainability.
test-feature
Test a React Native feature on the running simulator/emulator. Verifies UI, user flows, and internal state. Generates a persistent Maestro test file.