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/marcmallet/specforge/implementgit clone --depth 1 https://github.com/marcmallet/specforgeWhat 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 | $0.00015 | $0.00588 |
| Opus 5 | $0.00008 | $0.00294 |
| Sonnet 5 | $0.00003 | $0.00118 |
| Haiku 4.5 | $0.00002 | $0.00059 |
Grade A, and why
implement scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 66 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are coordinating the implementation of a feature using sub-agents.
Specs directory
Read the project config at: @.claude/specforge/config.json
If the file exists, use the value of the specsDir field as the base directory.
If the file does not exist, fall back to .claude/specforge/ inside the current project.
The full path for this ticket is: /$ARGUMENTS/
Read the spec at: /$ARGUMENTS/spec.md Read the plan at: /$ARGUMENTS/plan.md (if it does not exist, generate the plan first before implementing)
Agent Detection
Before spawning any agents, check for project-level agents in .claude/agents/:
- If
.claude/agents/backend.mdexists → use it for backend tasks, otherwise use the plugin'sagents/backend.md - If
.claude/agents/frontend.mdexists → use it for frontend tasks, otherwise use the plugin'sagents/frontend.md - If
.claude/agents/shared.mdexists → use it for shared tasks, otherwise use the plugin'sagents/shared.md
Phase 1 — Run in parallel
Spawn the backend and frontend agents at the same time. Do not wait for one before starting the other.
Backend Agent
Pass it:
- The full spec
- The full plan
- Only the tasks tagged [backend]
- The Shared Contracts section from the plan
Frontend Agent
Pass it:
- The full spec
- The full plan
- Only the tasks tagged [frontend]
- The Shared Contracts section from the plan
Phase 2 — Run after Phase 1 completes
Wait for BOTH backend and frontend agents to finish before starting.
Shared Agent
Pass it:
- The full spec
- The full plan
- Only the tasks tagged [both]
- The Shared Contracts section from the plan
- All files created or modified by the backend and frontend agents
After all agents finish
- Update the checklist in /$ARGUMENTS/plan.md marking all completed tasks as done
- If any agent reports an ambiguity, stop and ask the user before proceeding
- Do not implement anything outside the spec scope
- Summarise what was built, which agents were used (project or plugin), and list any deviations from the spec
- Remind the user to write tests for the implemented code
- Suggest the next step: "Ready to review? Run
/specforge:review $ARGUMENTS"
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 · 66 lines · 15 tokens per session scan A a1ea93789e13
implement is a command published in the GitHub repository marcmallet/specforge (9 stars, last pushed 4mo ago), licensed MIT. It adds 15 tokens to every session and 588 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.
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.