Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
56 changes: 15 additions & 41 deletions README.md
Original file line number Diff line number Diff line change
Expand Up @@ -53,52 +53,26 @@ python scripts/discover_registries.py --org dataciviclab --config dataciviclab.c
python scripts/discover_registries.py --org dataciviclab --config dataciviclab.config.yml --apply
```

## Tool MCP

Esposti via `agent-context-mcp` (server MCP `dataciviclab-context`).

| Tool | Quando usarlo |
|---|---|
| `session_bootstrap()` | Prima chiamata — orientamento rapido |
| `lab_map(domain=)` | Struttura del Lab: domains → repos → sources |
| `find(query, type=, domain=)` | Ricerca ranked su dataset/repo/domain/source/analysis |
| `explore(ref)` | Deep-dive di qualsiasi entità (slug repo, dataset, domain, source) |
| `workspace_triage(section=)` | Stato precisi: radar, prs, issues, registry, pipeline |
| `topic_index(resolve=)` | Legacy summary (resolve delega a explore) |
| `search(query)` | Legacy search (wraps find + GitHub issues) |
| `refresh_context()` | Trigger rebuild CI |

### Esempi

```python
# Orientamento
session_bootstrap()
# → markdown con radar, PR, issues, discussions

# Struttura del Lab
lab_map()
# → {domains: [...], repos: [...], sources: {...}, totals: {...}}

# Trovare un dataset
find("rifiuti", type="dataset")
# → results[{type, slug, name, parent_repo, parent_domain, score}]

# Contesto di un repo
explore("open-ispra")
# → {type: repo, role, domain, datasets: [...], sources: [...]}

# Stato radar
workspace_triage(section="radar")
# → {"green": 35, "yellow": 0, "red": 1, ...}
```
## Tool MCP — migrato a dcl-mcp

Il server MCP (`session_bootstrap`, `lab_map`, `find`, `explore`,
`workspace_triage`, `refresh_context`) non vive più qui: è migrato in
[`dataciviclab/dcl-mcp`](https://github.com/dataciviclab/dcl-mcp) insieme
ai tool Discussions e al nuovo `gcs_query`. ACB resta il **builder**:
CLI + CI che producono gli artifact sul branch `context`, che dcl-mcp
consuma da remoto.

Per l'uso degli artefatti via MCP, vedi il README di dcl-mcp. Le env
`ACB_REPO` / `ACB_BRANCH` / `ACB_ARTIFACT_TTL_SECONDS` continuano a
valere lì (puntano a questa repo come sorgente).

### Configurazione
### Configurazione (dcl-mcp)

```json
{
"mcpServers": {
"dataciviclab-context": {
"command": "agent-context-mcp",
"dataciviclab": {
"command": "dcl-mcp",
"env": { "GITHUB_TOKEN": "<opzionale>" }
}
}
Expand Down
3 changes: 1 addition & 2 deletions pyproject.toml
Original file line number Diff line number Diff line change
Expand Up @@ -18,7 +18,7 @@ dependencies = [
"click>=8.0",
"pydantic>=2.0",
"pyyaml>=6.0",
"lab-connectors[mcp] @ git+https://github.com/dataciviclab/lab-connectors.git",
"lab-connectors @ git+https://github.com/dataciviclab/lab-connectors.git",
]

[project.optional-dependencies]
Expand All @@ -32,7 +32,6 @@ dev = [

[project.scripts]
agent-context = "agent_context_builder.cli:main"
agent-context-mcp = "agent_context_builder.mcp_server:main"

[project.urls]
Repository = "https://github.com/dataciviclab/agent-context-builder"
Expand Down
Loading
Loading