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/hknc/claude-evolve/workflowgit clone --depth 1 https://github.com/hknc/claude-evolveWhat 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 | $0.00203 | $0.03009 |
| Opus 5 | $0.00102 | $0.01504 |
| Sonnet 5 | $0.00041 | $0.00602 |
| Haiku 4.5 | $0.00020 | $0.00301 |
Grade A, and why
workflow 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 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.
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 — 391 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/workflow Command
Workflow control using native Claude Code tasks. No persistent state files.
Subcommands
| Command | Action |
|---|---|
/workflow |
Show current task and phase (status) |
/workflow next |
Advance to next phase or sub-task |
/workflow skip |
Skip current phase or sub-task |
/workflow done |
Mark task complete (with confirmation) |
/workflow switch |
Switch between multiple active tasks |
/workflow expand |
Create sub-tasks from phase items |
Execution
1. Find Current Task
CALL TaskList tool to find tasks with [claude-evolve] prefix.
If multiple found:
- List them with index numbers
- Let user pick: "[claude-evolve] Multiple workflows found. Which one? (1-N)"
- Or suggest
/workflow switchfor explicit switching
If none found:
- Show fallback message (see below)
If one found:
- Use that task as current workflow
2. Show Status (default /workflow)
CALL TaskGet(id) to get full task details.
Read metadata to determine state:
metadata.depth- determines which phase sequence to usemetadata.phase- identifies current phasemetadata.phases_done- comma-separated list of completed phasesmetadata.phases_skipped- comma-separated list of skipped phases
Display format:
[claude-evolve] Workflow: {subject}
Phase: {current} | Progress: {visual progress bar}
{Phase-specific guidance based on current phase}
Commands: next | skip | done
Visual progress example for design depth:
Progress: requirements [OK] -> design [OK] -> implement <- testing
The <- indicates current phase.
3. Phase Transitions (/workflow next)
Advance to the next phase based on depth:
For design depth: requirements -> design -> implement -> testing -> done
For explore depth: research -> prototype -> evaluate -> decide -> done
For plan depth: action -> outcome -> done
Check for phase tasks with blockedBy:
- CALL TaskList to find
[claude-evolve] Phase N:tasks - If phase tasks exist with blockedBy, mark current phase task completed
- The next phase task automatically becomes available (no longer blocked)
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 · 391 lines · 203 tokens per session scan A d1c4b16f8efc
workflow is a command published in the GitHub repository hknc/claude-evolve (8 stars, last pushed 7mo ago), licensed MIT. It adds 203 tokens to every session and 3,009 once invoked, about $0.0010 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.
Other commands, from other repositories
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
analyze
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.
implement
Execute the implementation plan by processing and executing all tasks defined in tasks.md.