Configuring DevQL
DevQL is a thin client over the local Bitloops daemon. Run the repo-scoped commands below from inside a Git repository or Bitloops project.
Schema Bootstrap
bitloops devql init
The daemon bootstraps the DevQL schema automatically on startup. bitloops devql init remains available when you want to explicitly ensure the configured relational and event schemas exist.
Ingest Data
bitloops devql tasks enqueue --kind ingest
The CLI resolves repo policy locally, then sends ingestion requests to the daemon. Ingestion no longer owns schema bootstrap.
Sync Current State
bitloops devql tasks enqueue --kind sync
bitloops devql tasks enqueue --kind sync --status
bitloops devql tasks enqueue --kind sync --validate --status
bitloops devql tasks enqueue --kind sync now queues a sync task and returns immediately by default. Use --status when you want the CLI to follow that queued task until it completes or fails.
--validate queues a read-only validation task instead of mutating the current-state tables.
Successful sync tasks publish current-state generations. Built-in consumers such as test_harness.current_state and semantic_clones.current_state reconcile asynchronously from that feed, while historical ingest follow-up stays on the enrichment queue.
Query Data
bitloops devql query 'repo("bitloops")->artefacts(kind:"function")->limit(10)'
bitloops devql query '{ health { relational { backend connected } events { backend connected } } }'
Queries are DSL only when the input contains ->. Otherwise the CLI treats the input as raw GraphQL.
Semantic And Embedding Settings
Inference provider settings belong in the global daemon config:
[semantic_clones]
summary_mode = "auto"
embedding_mode = "semantic_aware_once"
ann_neighbors = 5
enrichment_workers = 1
[semantic_clones.inference]
summary_generation = "summary_llm"
code_embeddings = "local_code"
summary_embeddings = "local_code"
[inference.runtimes.bitloops_inference]
command = "/Users/alex/Library/Application Support/bitloops/tools/bitloops-inference/bitloops-inference"
args = []
startup_timeout_secs = 60
request_timeout_secs = 300
[inference.runtimes.bitloops_local_embeddings]
command = "/Users/alex/Library/Application Support/bitloops/tools/bitloops-local-embeddings/bitloops-local-embeddings"
args = []
startup_timeout_secs = 60
request_timeout_secs = 300
[inference.profiles.local_code]
task = "embeddings"
driver = "bitloops_embeddings_ipc"
runtime = "bitloops_local_embeddings"
model = "bge-m3"
cache_dir = "/Users/alex/.cache/bitloops-local-embeddings"
[inference.profiles.summary_llm]
task = "text_generation"
runtime = "bitloops_inference"
driver = "openai_chat_completions"
model = "gpt-5.4-mini"
api_key = "${OPENAI_API_KEY}"
base_url = "https://api.openai.com/v1/chat/completions"
temperature = "0.1"
max_output_tokens = 200
bitloops enable --install-embeddings and bitloops daemon enable --install-embeddings can create the default local embeddings profile for you. bitloops init --install-default-daemon can instead configure Bitloops cloud, the default local profile, or skip embeddings during onboarding. bitloops inference install manages the standalone summary runtime, and interactive bitloops enable or bitloops init --install-default-daemon can attach summaries to Ollama automatically when it is available. Edit the daemon config manually only when you want a hosted profile or a customised local profile.
When Bitloops installs a managed runtime, it writes an absolute path under the Bitloops data directory, as shown above. Use command = "bitloops-local-embeddings" or command = "bitloops-inference" only when you are managing those standalone binaries yourself on PATH.
Watch Behaviour
Watcher behaviour belongs in repo policy:
[watch]
watch_debounce_ms = 750
watch_poll_fallback_ms = 2500
Troubleshooting
bitloops status
bitloops devql packs --with-health
bitloops checkpoints status --detailed
bitloops --connection-status
Use bitloops status for daemon health, bitloops devql packs --with-health for capability-pack and embeddings health, and bitloops checkpoints status --detailed for policy root and fingerprint debugging.
bitloops status also shows sync queue totals, and when run inside a repository it includes the active or most recent sync task for that repo.