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/chris-dare-dev/agent-kit/roadmapgit clone --depth 1 https://github.com/chris-dare-dev/agent-kitWhat 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 | $0.00124 | $0.05252 |
| Opus 5 | $0.00062 | $0.02626 |
| Sonnet 5 | $0.00025 | $0.01050 |
| Haiku 4.5 | $0.00012 | $0.00525 |
Grade A, and why
roadmap 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 — 381 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Roadmap Orchestrator
You are the orchestrator. The main session IS the pipeline. Subagents cannot spawn subagents — every dispatch comes from here.
A four-phase gated roadmap builder. Auto-advances within a phase; gates between phases ONLY when there is an architecturally divergent fork the user must resolve. Always gates on external writes per workspace CLAUDE.md.
Phase 1 REFINE → roadmap-refiner → ## Goal section in plans/<slug>-roadmap.md
Phase 2 DECOMPOSE → roadmap-decomposer → ## Epics section
Phase 3 SEQUENCE → roadmap-sequencer → ## Roadmap — Now / Next / Later section
Phase 4 MATERIALIZE → roadmap-materializer → ## Cross-references + optional GitLab drafts
Parsing $ARGUMENTS
<slug-or-topic>— everything before the first--. Used verbatim as the brief if short and specific (≤40 chars, kebab-case → slug). If longer prose, treat as the brief and derive the slug from it (kebab-case, ≤40 chars).--gitlab— Phase 4 materializer drafts GitLab issue bodies. The orchestrator presents the drafts and waits for explicit user authorization before calling/issue-create.--iteration <id>— if--gitlaband issues are created, assign them to this iteration id.--resume— skip the init check; read the existing roadmap doc and jump to the first unpopulated section.
If $ARGUMENTS is empty, check whether the current conversation contains a problem being discussed. If yes, derive the brief from the conversation (see Step 0). If no problem context exists, STOP and ask: "What would you like to roadmap?"
When to invoke / When NOT to invoke
Invoke for:
- Multi-epic plans (≥ 2 epics spanning ≥ 2 weeks of work)
- When the user says "plan this project", "build a roadmap for X", "decompose this into milestones", "what's the path to ship Y", "help me sequence this work"
- Mid-conversation: the user has been discussing a problem and now wants a structured plan
Do NOT invoke for:
- Single milestone known and well-scoped → go straight to
/milestone-pipeline <id> - Single-issue ticket creation → use
/issue-create - Status or promotion check → use
/env-promotion-diff,/argocd-status - Re-plan of an already-shipped milestone → write
plans/<slug>-retrospective.mddirectly;/roadmapis forward-looking - No problem statement at all → STOP and ask what the user wants to roadmap
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 · 381 lines · 124 tokens per session scan A f3274f304d5b
roadmap is a command published in the GitHub repository chris-dare-dev/agent-kit (0 stars, last pushed 24d ago), licensed MIT. It adds 124 tokens to every session and 5,252 once invoked, about $0.0006 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.