ddb8a14ee9
* feat(skills): add a litellm-setup agent skill * fix(skills): run the quickstart for Docker with Postgres in litellm-setup Use scripts/quickstart.sh for the Docker plus Postgres path instead of a hand-rolled Postgres container, keep uv with the user's own Postgres, load .env as data instead of sourcing it, pin the uv Python range, chmod an existing .env, list bedrock_converse and vertex_ai rows in the cost map lookup, ask before editing the shell profile, and stop the health wait when the proxy exits * fix(skills): guide empty Postgres, docker sudo, and key fixes in litellm-setup Ask for a new empty database on the uv path and DIRECT_URL behind a pooler, add the baseline error to troubleshooting, and drop inherited database settings in the loader line. Check python3 and lsof, explain Docker's permission denied, give quickstart commands by the folder's path, and fix provider keys by recreating the container or deleting the quickstart model. Drop the structure-only test. * fix(skills): keep shell profiles untouched when litellm-setup installs uv The uv installer edits shell profiles unless UV_NO_MODIFY_PATH is set, so the skill now sets it and puts ~/.local/bin on PATH for later commands. It also says litellm --setup runs in a separate terminal and how to retire the shared master-key UI login * fix(skills): offer litellm --setup only on the uv path The Docker paths install no host litellm CLI, and the old wording also passed the skill's own step number on to the user * fix(skills): ask before using an env provider key, print model names config.yaml accepts, probe the port with curl * fix(skills): state the env key default in the question, chmod the wizard config, keep uv's bin dir on PATH * fix(skills): keep litellm-setup on its own .env and install when the shell has others * fix(skills): read the setup wizard's config without printing keys and hand off the exact restart command * fix(skills): hide the database-access dev skill from end-user skill installs * fix(skills): give the litellm --setup path a restart command that keeps its config * docs(skills): point litellm --setup users to the master key in litellm_config.yaml * fix(skills): hide every non-end-user skill, use Azure deployment names, keep the restart block out of the user's shell * test(skills): find non-end-user skills in-process instead of through git * docs(skills): price Azure deployments with base_model and put uv on PATH in the start block
2.5 KiB
| name | description | metadata | ||
|---|---|---|---|---|
| rust-tracing | Add or change Rust diagnostic tracing in litellm-rust, including route spans, subscriber layers, and Python logger delivery |
|
Rust tracing
Use upstream tracing throughout Rust, including #[tracing::instrument], events, and span propagation. Centralize collection and delivery infrastructure in crates/tracing. Direct upstream imports still reach our configured subscriber; re-exporting macros does not control delivery. Do not introduce Rust log or pyo3-log for this path
litellm-tracing owns shared subscriber layers, span field collection, and diagnostic processing. Keep adapters composable as tracing_subscriber::Layers, with Logger providing host setup. Runtime-specific delivery belongs in the host bridge. The Python bridge delivers directly to the existing Python SDK logger, preserving its handlers, filtering, redaction, and request correlation. Keep Python dependencies out of crates/tracing
Hosts configure subscribers. Keep Python execution scoped to its captured dispatch rather than installing a process-wide subscriber. Propagate both span context and dispatch across spawned work and returned streams
In core, instrument execution shared by native calls and hosted machines. Use consistent route, model, provider, streaming, and outcome fields. Put status recording at shared provider boundaries instead of scattering basic logging through handlers. Keep upstream HTTP status separate from route success
Use skip_all and explicitly selected fields. Basic tracing excludes bodies, credentials, headers, and raw error strings. Avoid automatic ret or err capture of sensitive values. Keep payload diagnostics separate and subject to existing redaction
A returned stream retains its route span until exhaustion, error, or drop, with exactly one terminal outcome. Builder construction does not start a trace. Never hold a span entry guard across an await. Diagnostic tracing remains separate from lifecycle callbacks and CustomLogger dispatch
Use litellm_tracing::sink_layer to compose a sink with other subscriber layers. It inherits span fields into events and emits span-close summaries with elapsed time. Test observable records, concurrent isolation, dynamic filtering, sensitive-field exclusion, and stream cancellation when changing this behavior
Consult the tracing API and subscriber layers for implementation details