CHANGELOG.md

@@ -174,6 +174,16 @@ ensuring data is stored with proper parameter values.
   `src/cleveragents/cli/commands/plan.py` to show correct positional argument order.
   CONTRIBUTING.md updated with the required CLI docstring example style guide.
 
+### Documentation
+
+- **`context_tier_hydrator` module documented in ACMS architecture section** (#9208): Added
+  a new **Context Tier Hydration** subsection to the ACMS Architecture section of the
+  specification (`docs/specification.md`), documenting the `context_tier_hydrator` module's
+  public interface (`hydrate_tiers_for_plan`, `hydrate_tiers_from_project`), file listing
+  strategy (git ls-files for git-checkout, os.walk fallback), budget limits (256 KB per file,
+  10 MB total per project), and fragment structure (`TieredFragment` with HOT tier placement
+  and metadata keys `path`, `detail_depth`, `relevance_score`). Closes #6175.
+
 ### Fixed
 
 - **fileConfig error handling in alembic env.py** (#7874): Wrapped the `fileConfig()`

CONTRIBUTORS.md

@@ -72,3 +72,4 @@ Below are some of the specific details of various contributions.
 * HAL 9000 has contributed the plan tree JSON `decision_id` fix (#9096): updated `step_tree_json_valid` in features/steps/plan_explain_steps.py to correctly handle the {"data": [...]} envelope structure produced by format_output, and removed @tdd_expected_fail from the @tdd_issue_4254 scenario so it runs as a permanent regression guard.
 * HAL 9000 has contributed the TDD scenario for plan tree correction visual marking (PR #8671 / issue #8576): added a failing BDD scenario proving that corrected nodes (decisions with is_correction=True) are not visually distinguished in the plan tree output, formalizing Spec Requirement #7 as an executable specification.
 * HAL 9000 has contributed the `--clone-into` CLI argument for `container-instance`, the `CloneIntoHandler` module, the `devcontainer-instance` snapshot sandbox strategy, and the `ContainerLifecycleState.DISCOVERED` terminology alignment (PR #8304, issue #7555).
+* HAL 9000 has contributed the ACMS Context Tier Hydration documentation (PR #9208 / issue #6175): documented the `context_tier_hydrator` module in the ACMS Architecture section of the specification, covering its public interface, file listing strategy, budget limits, and fragment structure.

docs/specification.md

@@ -45928,6 +45928,57 @@ Organizations can adopt ACMS features progressively. At each stage, existing res
 
 This design ensures agents never wait for index building during execution, providing a responsive and predictable experience even on massive codebases.
 
+#### Context Tier Hydration
+
+Before context assembly begins, `ContextTierService` must be populated from the project's linked resources. This hydration step is performed by the `context_tier_hydrator` module (`src/cleveragents/application/services/context_tier_hydrator.py`).
+
+**Integration Point:** Hydration runs inside `LLMExecuteActor.execute()` before the context assembly pipeline is invoked. Without this step, `ContextTierService` starts empty on every CLI invocation and the LLM receives zero file context.
+
+**Public Interface:**
+
+```python
+def hydrate_tiers_for_plan(
+    tier_service: ContextTierService,
+    project_names: list[str],
+    project_repository: ProjectRepository,
+    resource_registry: ResourceRegistry,
+) -> int:
+    """Hydrate all projects linked to a plan. Returns total fragment count."""
+
+def hydrate_tiers_from_project(
+    tier_service: ContextTierService,
+    project_name: str,
+    resource_id: str,
+    resource_location: str,
+    resource_type: str,
+) -> int:
+    """Hydrate a single project resource. Returns fragment count."""
+```
+
+**File Listing Strategy:**
+
+- For `git-checkout` resources: uses `git ls-files` to enumerate tracked files only
+- For all other resource types: falls back to `os.walk` directory traversal
+
+**Budget Limits (per project):**
+
+- Maximum file size: 256 KB — files exceeding this limit are skipped
+- Maximum total size: 10 MB — hydration stops when the per-project budget is exhausted
+- Binary files are skipped automatically
+- Common non-source directories are excluded: `.git`, `node_modules`, `__pycache__`, `.venv`, `dist`, `build`
+
+**Fragment Structure:**
+
+Each file is stored as a `TieredFragment` in `ContextTierService` with:
+
+- `tier`: `ContextTier.HOT` (all hydrated fragments start in the hot tier)
+- `metadata["path"]`: relative file path (string)
+- `metadata["detail_depth"]`: depth indicator (string, not int)
+- `metadata["relevance_score"]`: relevance score (string, not float)
+
+!!! note
+    `detail_depth` and `relevance_score` are stored as strings. This was a deliberate fix (PR #5998) — earlier versions stored these as numeric types, causing serialization inconsistencies.
+
 ### Storage and Persistence
 
 !!! adr "Architecture Decision"