@thatix.io/context-first-agents-cli 0.2.1 → 0.2.2

package/dist/templates/commands/en/orchestrate.md

@@ -93,6 +93,39 @@ See `agents/CONTEXT-CONTRACT.md` for the exact shape. In short:
 - **limits**: `contextPolicy` (default `select-do-not-dump`), `maxFilesPerWorker`
 - **return**: summary, changes, evidence, tests, unresolved questions, confidence
 
+## Step 5b — Prepare the session worktrees (via git, not Node)
+
+Before spawning any agent, create an **isolated git worktree per impacted repository**
+(only the ones in the graph), so each implementer has a place to write without touching
+the main repo. Use `base_path` (from `ai.properties.md`) and the `<ISSUE-ID>`.
+
+For each impacted repository `<repo>`:
+
+1. If `.sessions/<ISSUE-ID>/<repo>/` already exists, **skip** (worktree ready).
+2. Check whether branch `feature/<ISSUE-ID>` already exists in the repo:
+   ```bash
+   git -C "{base_path}/<repo>" rev-parse --verify --quiet "feature/<ISSUE-ID>"
+   ```
+3. Create the worktree from the main repo:
+   - if the branch does **not** exist (create it in the worktree):
+     ```bash
+     git -C "{base_path}/<repo>" worktree add -b "feature/<ISSUE-ID>" \
+         "$(pwd)/.sessions/<ISSUE-ID>/<repo>"
+     ```
+   - if the branch **already** exists (reuse it):
+     ```bash
+     git -C "{base_path}/<repo>" worktree add \
+         "$(pwd)/.sessions/<ISSUE-ID>/<repo>" "feature/<ISSUE-ID>"
+     ```
+
+Rules:
+- **Never** `checkout` in the main repo (`{base_path}/<repo>`) — the worktree isolates everything.
+- If `git worktree add` fails with "already exists", treat it as ready and continue.
+- Only prepare worktrees for the **impacted** repos in the graph, not every repo in the manifest.
+- Record in `execution-plan.md` which worktrees were created (path + branch).
+
+After this, each agent's `writeBoundary` (`.sessions/<ISSUE-ID>/<repo>/`) actually exists.
+
 ## Step 6 — Spawn ephemeral agents (Task tool)
 
 Execute the DAG respecting `dependsOn`:

package/dist/templates/commands/es/orchestrate.md

@@ -94,6 +94,39 @@ Ver `agents/CONTEXT-CONTRACT.md` para el formato exacto. En resumen:
 - **limits**: `contextPolicy` (por defecto `select-do-not-dump`), `maxFilesPerWorker`
 - **return**: summary, changes, evidence, tests, unresolved, confidence
 
+## Paso 5b — Preparar los worktrees de la sesión (vía git, no Node)
+
+Antes de spawnear cualquier agente, crea un **git worktree aislado por repositorio
+impactado** (sólo los del grafo), para que cada implementer tenga dónde escribir sin tocar
+el repo principal. Usa `base_path` (de `ai.properties.md`) y el `<ISSUE-ID>`.
+
+Para cada repositorio impactado `<repo>`:
+
+1. Si `.sessions/<ISSUE-ID>/<repo>/` ya existe, **sáltalo** (worktree listo).
+2. Chequea si la branch `feature/<ISSUE-ID>` ya existe en el repo:
+   ```bash
+   git -C "{base_path}/<repo>" rev-parse --verify --quiet "feature/<ISSUE-ID>"
+   ```
+3. Crea el worktree desde el repo principal:
+   - si la branch **no** existe (créala en el worktree):
+     ```bash
+     git -C "{base_path}/<repo>" worktree add -b "feature/<ISSUE-ID>" \
+         "$(pwd)/.sessions/<ISSUE-ID>/<repo>"
+     ```
+   - si la branch **ya** existe (reutilízala):
+     ```bash
+     git -C "{base_path}/<repo>" worktree add \
+         "$(pwd)/.sessions/<ISSUE-ID>/<repo>" "feature/<ISSUE-ID>"
+     ```
+
+Reglas:
+- **Nunca** hagas `checkout` en el repo principal (`{base_path}/<repo>`) — el worktree aísla todo.
+- Si `git worktree add` falla con "already exists", trátalo como listo y continúa.
+- Sólo prepara worktrees de los repos **impactados** del grafo, no de todos los del manifiesto.
+- Registra en `execution-plan.md` qué worktrees se crearon (path + branch).
+
+Tras esto, el `writeBoundary` de cada agente (`.sessions/<ISSUE-ID>/<repo>/`) existe de verdad.
+
 ## Paso 6 — Spawnear los agentes efímeros (Task tool)
 
 Ejecuta el DAG respetando `dependsOn`:

package/dist/templates/commands/pt-BR/orchestrate.md

@@ -94,6 +94,39 @@ Veja `agents/CONTEXT-CONTRACT.md` para o formato exato. Em resumo:
 - **limits**: `contextPolicy` (padrão `select-do-not-dump`), `maxFilesPerWorker`
 - **return**: summary, changes, evidence, tests, unresolved, confidence
 
+## Passo 5b — Preparar os worktrees da sessão (via git, não Node)
+
+Antes de spawnar qualquer agente, crie um **git worktree isolado por repositório
+impactado** (só os do grafo), para que cada implementer tenha onde escrever sem tocar no
+repo principal. Use `base_path` (de `ai.properties.md`) e o `<ISSUE-ID>`.
+
+Para cada repositório impactado `<repo>`:
+
+1. Se `.sessions/<ISSUE-ID>/<repo>/` já existir, **pule** (worktree já preparado).
+2. Descubra se a branch `feature/<ISSUE-ID>` já existe no repo:
+   ```bash
+   git -C "{base_path}/<repo>" rev-parse --verify --quiet "feature/<ISSUE-ID>"
+   ```
+3. Crie o worktree a partir do repo principal:
+   - se a branch **não** existe (cria a branch no worktree):
+     ```bash
+     git -C "{base_path}/<repo>" worktree add -b "feature/<ISSUE-ID>" \
+         "$(pwd)/.sessions/<ISSUE-ID>/<repo>"
+     ```
+   - se a branch **já** existe (reaproveita):
+     ```bash
+     git -C "{base_path}/<repo>" worktree add \
+         "$(pwd)/.sessions/<ISSUE-ID>/<repo>" "feature/<ISSUE-ID>"
+     ```
+
+Regras:
+- **Nunca** faça `checkout` no repo principal (`{base_path}/<repo>`) — o worktree isola tudo.
+- Se `git worktree add` falhar por "already exists", trate como já preparado e siga.
+- Só prepare worktrees dos repos **impactados** pelo grafo, não de todos do manifesto.
+- Registre no `execution-plan.md` quais worktrees foram criados (path + branch).
+
+Depois disso, o `writeBoundary` de cada agente (`.sessions/<ISSUE-ID>/<repo>/`) existe de fato.
+
 ## Passo 6 — Spawnar os agentes efêmeros (Task tool)
 
 Execute o DAG respeitando `dependsOn`:

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@thatix.io/context-first-agents-cli",
-  "version": "0.2.1",
+  "version": "0.2.2",
   "publishConfig": {
     "access": "public"
   },

package/templates/commands/en/orchestrate.md

@@ -93,6 +93,39 @@ See `agents/CONTEXT-CONTRACT.md` for the exact shape. In short:
 - **limits**: `contextPolicy` (default `select-do-not-dump`), `maxFilesPerWorker`
 - **return**: summary, changes, evidence, tests, unresolved questions, confidence
 
+## Step 5b — Prepare the session worktrees (via git, not Node)
+
+Before spawning any agent, create an **isolated git worktree per impacted repository**
+(only the ones in the graph), so each implementer has a place to write without touching
+the main repo. Use `base_path` (from `ai.properties.md`) and the `<ISSUE-ID>`.
+
+For each impacted repository `<repo>`:
+
+1. If `.sessions/<ISSUE-ID>/<repo>/` already exists, **skip** (worktree ready).
+2. Check whether branch `feature/<ISSUE-ID>` already exists in the repo:
+   ```bash
+   git -C "{base_path}/<repo>" rev-parse --verify --quiet "feature/<ISSUE-ID>"
+   ```
+3. Create the worktree from the main repo:
+   - if the branch does **not** exist (create it in the worktree):
+     ```bash
+     git -C "{base_path}/<repo>" worktree add -b "feature/<ISSUE-ID>" \
+         "$(pwd)/.sessions/<ISSUE-ID>/<repo>"
+     ```
+   - if the branch **already** exists (reuse it):
+     ```bash
+     git -C "{base_path}/<repo>" worktree add \
+         "$(pwd)/.sessions/<ISSUE-ID>/<repo>" "feature/<ISSUE-ID>"
+     ```
+
+Rules:
+- **Never** `checkout` in the main repo (`{base_path}/<repo>`) — the worktree isolates everything.
+- If `git worktree add` fails with "already exists", treat it as ready and continue.
+- Only prepare worktrees for the **impacted** repos in the graph, not every repo in the manifest.
+- Record in `execution-plan.md` which worktrees were created (path + branch).
+
+After this, each agent's `writeBoundary` (`.sessions/<ISSUE-ID>/<repo>/`) actually exists.
+
 ## Step 6 — Spawn ephemeral agents (Task tool)
 
 Execute the DAG respecting `dependsOn`:

package/templates/commands/es/orchestrate.md

@@ -94,6 +94,39 @@ Ver `agents/CONTEXT-CONTRACT.md` para el formato exacto. En resumen:
 - **limits**: `contextPolicy` (por defecto `select-do-not-dump`), `maxFilesPerWorker`
 - **return**: summary, changes, evidence, tests, unresolved, confidence
 
+## Paso 5b — Preparar los worktrees de la sesión (vía git, no Node)
+
+Antes de spawnear cualquier agente, crea un **git worktree aislado por repositorio
+impactado** (sólo los del grafo), para que cada implementer tenga dónde escribir sin tocar
+el repo principal. Usa `base_path` (de `ai.properties.md`) y el `<ISSUE-ID>`.
+
+Para cada repositorio impactado `<repo>`:
+
+1. Si `.sessions/<ISSUE-ID>/<repo>/` ya existe, **sáltalo** (worktree listo).
+2. Chequea si la branch `feature/<ISSUE-ID>` ya existe en el repo:
+   ```bash
+   git -C "{base_path}/<repo>" rev-parse --verify --quiet "feature/<ISSUE-ID>"
+   ```
+3. Crea el worktree desde el repo principal:
+   - si la branch **no** existe (créala en el worktree):
+     ```bash
+     git -C "{base_path}/<repo>" worktree add -b "feature/<ISSUE-ID>" \
+         "$(pwd)/.sessions/<ISSUE-ID>/<repo>"
+     ```
+   - si la branch **ya** existe (reutilízala):
+     ```bash
+     git -C "{base_path}/<repo>" worktree add \
+         "$(pwd)/.sessions/<ISSUE-ID>/<repo>" "feature/<ISSUE-ID>"
+     ```
+
+Reglas:
+- **Nunca** hagas `checkout` en el repo principal (`{base_path}/<repo>`) — el worktree aísla todo.
+- Si `git worktree add` falla con "already exists", trátalo como listo y continúa.
+- Sólo prepara worktrees de los repos **impactados** del grafo, no de todos los del manifiesto.
+- Registra en `execution-plan.md` qué worktrees se crearon (path + branch).
+
+Tras esto, el `writeBoundary` de cada agente (`.sessions/<ISSUE-ID>/<repo>/`) existe de verdad.
+
 ## Paso 6 — Spawnear los agentes efímeros (Task tool)
 
 Ejecuta el DAG respetando `dependsOn`:

package/templates/commands/pt-BR/orchestrate.md

@@ -94,6 +94,39 @@ Veja `agents/CONTEXT-CONTRACT.md` para o formato exato. Em resumo:
 - **limits**: `contextPolicy` (padrão `select-do-not-dump`), `maxFilesPerWorker`
 - **return**: summary, changes, evidence, tests, unresolved, confidence
 
+## Passo 5b — Preparar os worktrees da sessão (via git, não Node)
+
+Antes de spawnar qualquer agente, crie um **git worktree isolado por repositório
+impactado** (só os do grafo), para que cada implementer tenha onde escrever sem tocar no
+repo principal. Use `base_path` (de `ai.properties.md`) e o `<ISSUE-ID>`.
+
+Para cada repositório impactado `<repo>`:
+
+1. Se `.sessions/<ISSUE-ID>/<repo>/` já existir, **pule** (worktree já preparado).
+2. Descubra se a branch `feature/<ISSUE-ID>` já existe no repo:
+   ```bash
+   git -C "{base_path}/<repo>" rev-parse --verify --quiet "feature/<ISSUE-ID>"
+   ```
+3. Crie o worktree a partir do repo principal:
+   - se a branch **não** existe (cria a branch no worktree):
+     ```bash
+     git -C "{base_path}/<repo>" worktree add -b "feature/<ISSUE-ID>" \
+         "$(pwd)/.sessions/<ISSUE-ID>/<repo>"
+     ```
+   - se a branch **já** existe (reaproveita):
+     ```bash
+     git -C "{base_path}/<repo>" worktree add \
+         "$(pwd)/.sessions/<ISSUE-ID>/<repo>" "feature/<ISSUE-ID>"
+     ```
+
+Regras:
+- **Nunca** faça `checkout` no repo principal (`{base_path}/<repo>`) — o worktree isola tudo.
+- Se `git worktree add` falhar por "already exists", trate como já preparado e siga.
+- Só prepare worktrees dos repos **impactados** pelo grafo, não de todos do manifesto.
+- Registre no `execution-plan.md` quais worktrees foram criados (path + branch).
+
+Depois disso, o `writeBoundary` de cada agente (`.sessions/<ISSUE-ID>/<repo>/`) existe de fato.
+
 ## Passo 6 — Spawnar os agentes efêmeros (Task tool)
 
 Execute o DAG respeitando `dependsOn`: