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/nicelight/devrails-26/map-codebasegit clone --depth 1 https://github.com/nicelight/DevRails-26What 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 | $0.00019 | $0.00835 |
| Opus 5 | $0.00010 | $0.00417 |
| Sonnet 5 | $0.00004 | $0.00167 |
| Haiku 4.5 | $0.00002 | $0.00084 |
Grade A, and why
map-codebase 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.
What it actually says
/map-codebase — Brownfield mapping
- Создай
.tasks/TASK-MB-MAP/. - Выбери минимально достаточную тактику discovery по фактическому размеру и структуре repository:
- для малого repository или очевидного полного read set используй direct reads одним агентом;
- для действительно широкого discovery
/context-manifestили bounded delegation допустимы только когда это дешевле direct reads и текущие роль и operator contract разрешают delegation; - delegation не является default; delegated agent не запускает других агентов;
- поисковые globs должны матчиться, а evidence должно покрывать применимые code, config, CI/tooling, data/state и test surfaces.
Записывай substantive mapping evidence и отчёты в .tasks/TASK-MB-MAP/
независимо от выбранной тактики.
- Синтезируй evidence-backed as-is current-state baseline
.memory-bank/по чеклисту:
- product
- architecture (C4)
- spec-index / spec-backbone / glossary / invariants (только для регистрации существующих artifacts и явно обозначенного current-state routing; не выводи normative target или Global Backbone decision из кода)
- runbooks
- contracts
- states (если lifecycle/state rules очевидны из кода, workflow или тестов)
- testing
- index
/map-codebase фиксирует change units/code roots, imports/calls,
events/background work, writers/state paths, runtime entrypoints и proof paths
только как as-is evidence, не target authority или accepted graph rows.
PRD-less rule (non-negotiable): если нет
prd.md, запрещено генерировать roadmap сущности:
.memory-bank/epics/*.memory-bank/features/*.memory-bank/tasks/*.task.jsontask records.memory-bank/tasks/index.jsontask linksМаппинг = as-is документация по evidence, а не планирование.
- Проверь качество baseline:
- отдели facts от inferences и явно привяжи их к evidence;
- противоречия, недоказанные выводы и неполное покрытие запиши как
needs verification/ unresolved evidence; - не заменяй существующие baseline facts догадками.
- помечай as-is claims как current state, чтобы downstream skill не принял их за желаемую архитектуру;
- если обновляемый artifact уже содержит accepted target, сохрани его и явно
отдели от current state; не создавай и не изменяй target
AD-*, normative rules или architecture decisions на основании одного as-is evidence; - если current отличается от accepted target, зафиксируй оба состояния и уже известный reconciliation route в существующем owning artifact. Само отличие является delta, а не authority conflict; неизвестную material route оставь unresolved для её downstream owner. Не создавай второй dependency graph или обязательный artifact для этого разделения.
- Заверши immediate handoff:
- если authoritative PRD/delta уже передан вызывающим workflow, сохрани его как downstream input и не спрашивай повторно; верни invoking workflow/current PRD owner ссылки на baseline вместе с тем же переданным PRD/delta;
- если PRD/delta не передан, заверши as-is baseline, запроси его у оператора и остановись;
- не воспроизводи дальнюю downstream chain и не переходи к roadmap, task
generation или execution из
/map-codebase.
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 · 80 lines · 19 tokens per session scan A 737784437a1b
map-codebase is a command published in the GitHub repository nicelight/DevRails-26 (22 stars, last pushed 4d ago), licensed MIT. It adds 19 tokens to every session and 835 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
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.