Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
41 commits
Select commit Hold shift + click to select a range
486f323
ci: bump linear/linear-release-action from 0.18.1 to 0.18.2 (#261)
dependabot[bot] Nov 6, 2026
efe0cfd
chore(deps): bump mypy from 2.4.0 to 2.4.1 (#263)
dependabot[bot] Nov 6, 2026
9b5ac56
chore(deps): bump ruff from 0.17.2 to 0.17.3 (#264)
dependabot[bot] Nov 6, 2026
afdeddf
ci: bump astral-sh/setup-uv from 10.3.2 to 10.3.3 (#262)
dependabot[bot] Nov 6, 2026
8897659
chore(deps): bump litellm from 1.106.2 to 1.106.6 (#265)
dependabot[bot] Nov 6, 2026
60ca6a6
chore(deps): bump openinference-instrumentation-llama-index from 4.6.…
dependabot[bot] Nov 6, 2026
ebee042
chore(deps): bump openinference-instrumentation-anthropic from 3.0.3 …
dependabot[bot] Nov 6, 2026
fa036c1
fix: cover langchain 0.1.82 and llama-index 4.7.2 in the span-context…
PabloPardoGarcia Nov 6, 2026
cc12396
chore(main): release 1.5.6 (#258)
github-actions[bot] Nov 6, 2026
fa24f52
chore(deps): bump opentelemetry-exporter-otlp-proto-http from 1.45.1 …
dependabot[bot] Nov 9, 2026
8143b18
chore(deps): bump langchain-openai from 1.7.2 to 1.7.4 (#271)
dependabot[bot] Nov 9, 2026
d084f39
chore(deps): bump openinference-instrumentation-anthropic from 3.1.0 …
dependabot[bot] Nov 9, 2026
56c6e75
chore(deps): bump opentelemetry-api from 1.45.1 to 1.45.2 (#273)
dependabot[bot] Nov 9, 2026
4eb2dd3
test: count abandoned-stream usage per call, not across calls (#275)
PabloPardoGarcia Nov 9, 2026
96a56b0
chore(deps): bump openinference-instrumentation-openai from 0.1.65 to…
dependabot[bot] Nov 9, 2026
bd240cb
chore: identical-argument scenario for the hints harness (RIUS-1727)
baharedr Nov 12, 2026
09d2ca0
chore: the hints harness sets each run's answer as its root output
baharedr Nov 12, 2026
bcace1f
chore: hints harness tasks that never ask the model to repeat a call …
baharedr Nov 12, 2026
5166214
chore: hints harness summaries carry the model calls' token usage
baharedr Nov 12, 2026
125ffca
chore: put the token counts in the hints harness summary line
baharedr Nov 12, 2026
643f536
chore: deterministic hints harness faults for the measurement's negat…
baharedr Nov 12, 2026
19d4eaf
fix: cover langchain 0.1.86 and llama-index 4.7.6 in the span-context…
PabloPardoGarcia Nov 16, 2026
3e1be84
chore(deps): bump pytest from 9.1.1 to 9.2.0 (#277)
dependabot[bot] Nov 16, 2026
f0f00e8
chore(deps): bump anthropic from 2.1.0 to 2.8.0 (#278)
dependabot[bot] Nov 16, 2026
8d11058
chore(deps): bump openinference-instrumentation-anthropic from 3.3.0 …
dependabot[bot] Nov 16, 2026
e1d381e
chore(deps): bump openinference-instrumentation-llama-index from 4.6.…
dependabot[bot] Nov 16, 2026
6160925
chore(deps): bump litellm from 1.106.6 to 1.109.1 (#279)
dependabot[bot] Nov 16, 2026
d9ba735
chore(main): release 1.5.7 (#283)
github-actions[bot] Nov 16, 2026
8f954b9
fix: land the 2026-11-16 GenAI semconv sweep and sync rius-semconv v1…
PabloPardoGarcia Nov 16, 2026
e46f3d1
perf: step over JSON strings in the media rebuild walk instead of dec…
PabloPardoGarcia Nov 16, 2026
db59afc
fix: cover langchain 0.1.87, llama-index 4.7.7 and anthropic 3.3.6 in…
PabloPardoGarcia Nov 17, 2026
5741a2a
fix: map LiteLLM's ollama_chat provider to the registry's ollama (#292)
PabloPardoGarcia Nov 17, 2026
4df9784
fix: require openinference-instrumentation-anthropic 3.1.1 (#293)
PabloPardoGarcia Nov 17, 2026
c40cdf1
fix: drop the llm.tools copies only when spelled like their schema ke…
PabloPardoGarcia Nov 17, 2026
fef9fad
fix: retriever spans use OpenTelemetry span kind INTERNAL (#296)
PabloPardoGarcia Nov 17, 2026
6b54cdf
chore(main): release 1.6.0 (#290)
github-actions[bot] Nov 17, 2026
82f3de1
chore: hints harness tasks a real model may loop on unprompted: polli…
baharedr Nov 17, 2026
6a122d8
Merge remote-tracking branch 'origin/main' into rius-1727-identical-args
baharedr Nov 17, 2026
1804d83
fix(harness): grade structured task answers exactly
baharedr Nov 17, 2026
f0f8699
chore(harness): a refund case with work left after the status loop, a…
baharedr Nov 17, 2026
ec83185
chore(harness): three behavioural tasks, the deployed ping interval, …
baharedr Nov 18, 2026
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
12 changes: 6 additions & 6 deletions .github/workflows/ci.yml
Original file line number Diff line number Diff line change
Expand Up @@ -17,7 +17,7 @@ jobs:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
- uses: astral-sh/setup-uv@07a047e608db84b3d8c06cd9a7c7dcc9c355c18e # v10.3.2
- uses: astral-sh/setup-uv@68559e99cc807c1eb7e9238fd43a15bba460eec2 # v10.3.3
with:
enable-cache: true
- run: uv sync --locked --group dev
Expand All @@ -34,7 +34,7 @@ jobs:
python-version: ["3.10", "3.11", "3.12", "3.13", "3.14"]
steps:
- uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
- uses: astral-sh/setup-uv@07a047e608db84b3d8c06cd9a7c7dcc9c355c18e # v10.3.2
- uses: astral-sh/setup-uv@68559e99cc807c1eb7e9238fd43a15bba460eec2 # v10.3.3
with:
enable-cache: true
python-version: ${{ matrix.python-version }}
Expand All @@ -58,7 +58,7 @@ jobs:
RIUS_EXPECT_MCP_MAJOR: "1"
steps:
- uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
- uses: astral-sh/setup-uv@07a047e608db84b3d8c06cd9a7c7dcc9c355c18e # v10.3.2
- uses: astral-sh/setup-uv@68559e99cc807c1eb7e9238fd43a15bba460eec2 # v10.3.3
with:
enable-cache: true
- run: uv sync --locked --group dev
Expand All @@ -80,7 +80,7 @@ jobs:
RIUS_EXPECT_OPENAI_MAJOR: "3"
steps:
- uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
- uses: astral-sh/setup-uv@07a047e608db84b3d8c06cd9a7c7dcc9c355c18e # v10.3.2
- uses: astral-sh/setup-uv@68559e99cc807c1eb7e9238fd43a15bba460eec2 # v10.3.3
with:
enable-cache: true
- run: uv sync --locked --group dev
Expand Down Expand Up @@ -111,7 +111,7 @@ jobs:
--with setuptools<81
steps:
- uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
- uses: astral-sh/setup-uv@07a047e608db84b3d8c06cd9a7c7dcc9c355c18e # v10.3.2
- uses: astral-sh/setup-uv@68559e99cc807c1eb7e9238fd43a15bba460eec2 # v10.3.3
with:
enable-cache: true
python-version: "3.10"
Expand All @@ -129,7 +129,7 @@ jobs:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
- uses: astral-sh/setup-uv@07a047e608db84b3d8c06cd9a7c7dcc9c355c18e # v10.3.2
- uses: astral-sh/setup-uv@68559e99cc807c1eb7e9238fd43a15bba460eec2 # v10.3.3
with:
enable-cache: true
- run: uv sync --locked --group dev --all-extras
Expand Down
4 changes: 2 additions & 2 deletions .github/workflows/release.yml
Original file line number Diff line number Diff line change
Expand Up @@ -34,7 +34,7 @@ jobs:
contents: read
steps:
- uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
- uses: astral-sh/setup-uv@07a047e608db84b3d8c06cd9a7c7dcc9c355c18e # v10.3.2
- uses: astral-sh/setup-uv@68559e99cc807c1eb7e9238fd43a15bba460eec2 # v10.3.3
with:
enable-cache: true
- run: uv sync --locked --group dev
Expand Down Expand Up @@ -138,7 +138,7 @@ jobs:
if: ${{ steps.key.outputs.present == 'true' }}
with:
fetch-depth: 0 # the sync scans history back to the last recorded release
- uses: linear/linear-release-action@30f9ae77461ec29f07fffe0c52edd1909bfbb6f5 # v0.18.1
- uses: linear/linear-release-action@dd4fc407f860b9460d2cf0b3fb1edd8985cf443f # v0.18.2
if: ${{ steps.key.outputs.present == 'true' }}
with:
access_key: ${{ secrets.LINEAR_ACCESS_KEY }}
Expand Down
2 changes: 1 addition & 1 deletion .github/workflows/upstream-stopgaps.yml
Original file line number Diff line number Diff line change
Expand Up @@ -58,7 +58,7 @@ jobs:
timeout-minutes: 15
steps:
- uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
- uses: astral-sh/setup-uv@07a047e608db84b3d8c06cd9a7c7dcc9c355c18e # v10.3.2
- uses: astral-sh/setup-uv@68559e99cc807c1eb7e9238fd43a15bba460eec2 # v10.3.3
with:
enable-cache: true
# Re-resolve only the instrumentors and the libraries they hook
Expand Down
2 changes: 1 addition & 1 deletion .release-please-manifest.json
Original file line number Diff line number Diff line change
@@ -1,3 +1,3 @@
{
".": "1.5.5"
".": "1.6.0"
}
43 changes: 43 additions & 0 deletions CHANGELOG.md
Original file line number Diff line number Diff line change
@@ -1,5 +1,48 @@
# Changelog

## [1.6.0](https://github.com/glassflow/rius-sdk-python/compare/v1.5.7...v1.6.0) (2026-11-17)


### Features

* honour OTEL_INSTRUMENTATION_GENAI_CAPTURE_CONTENT when RIUS_CAPTURE_CONTENT is unset ([8f954b9](https://github.com/glassflow/rius-sdk-python/commit/8f954b905bf7a826a258f81cdd6a4fce60092653))
* map invoke_runner and other framework entrypoint operations to the CHAIN kind ([8f954b9](https://github.com/glassflow/rius-sdk-python/commit/8f954b905bf7a826a258f81cdd6a4fce60092653))


### Bug Fixes

* accept structured content attributes from third-party spans ([8f954b9](https://github.com/glassflow/rius-sdk-python/commit/8f954b905bf7a826a258f81cdd6a4fce60092653))
* cover langchain 0.1.87, llama-index 4.7.7 and anthropic 3.3.6 in the newest-instrumentor checks ([#294](https://github.com/glassflow/rius-sdk-python/issues/294)) ([db59afc](https://github.com/glassflow/rius-sdk-python/commit/db59afcaaff08ee40b6d2a0079badc022be860a5))
* drop MCP credentials from recorded tool definitions ([8f954b9](https://github.com/glassflow/rius-sdk-python/commit/8f954b905bf7a826a258f81cdd6a4fce60092653))
* drop the llm.tools copies only when spelled like their schema key, and cover the orphan and native cases ([#295](https://github.com/glassflow/rius-sdk-python/issues/295)) ([c40cdf1](https://github.com/glassflow/rius-sdk-python/commit/c40cdf1327a7b20d9fc1c8da9c630b0ebedce06c))
* map LiteLLM's ollama_chat provider to the registry's ollama ([#292](https://github.com/glassflow/rius-sdk-python/issues/292)) ([5741a2a](https://github.com/glassflow/rius-sdk-python/commit/5741a2afa8dbcdaaef331ca150369ef247112f98))
* record tool definitions in the conventions' function-tool shape ([8f954b9](https://github.com/glassflow/rius-sdk-python/commit/8f954b905bf7a826a258f81cdd6a4fce60092653))
* require openinference-instrumentation-anthropic 3.1.1 ([#293](https://github.com/glassflow/rius-sdk-python/issues/293)) ([4df9784](https://github.com/glassflow/rius-sdk-python/commit/4df9784a1f2045bb34571fa0ca2b7d2f592c1b84))
* retriever spans use OpenTelemetry span kind INTERNAL ([#296](https://github.com/glassflow/rius-sdk-python/issues/296)) ([fef9fad](https://github.com/glassflow/rius-sdk-python/commit/fef9fad818ad5973eb9122f67f11ae1c3f544e14))
* size a blob whose content is a complete strip replacement from that replacement ([8f954b9](https://github.com/glassflow/rius-sdk-python/commit/8f954b905bf7a826a258f81cdd6a4fce60092653))
* sync rius-semconv v1.13.0 ([8f954b9](https://github.com/glassflow/rius-sdk-python/commit/8f954b905bf7a826a258f81cdd6a4fce60092653))
* write a tool call without a name as a text part instead of an invalid tool-call part ([8f954b9](https://github.com/glassflow/rius-sdk-python/commit/8f954b905bf7a826a258f81cdd6a4fce60092653))


### Performance Improvements

* step over JSON strings in the media rebuild walk instead of decoding them ([#291](https://github.com/glassflow/rius-sdk-python/issues/291)) ([e46f3d1](https://github.com/glassflow/rius-sdk-python/commit/e46f3d100801a87dd756dab5bea36b0cc959f30f))

## [1.5.7](https://github.com/glassflow/rius-sdk-python/compare/v1.5.6...v1.5.7) (2026-11-16)


### Bug Fixes

* cover langchain 0.1.86 and llama-index 4.7.6 in the span-context stopgaps ([#282](https://github.com/glassflow/rius-sdk-python/issues/282)) ([19d4eaf](https://github.com/glassflow/rius-sdk-python/commit/19d4eaf87c3d2f6800469899f4f097d5506f8554))

## [1.5.6](https://github.com/glassflow/rius-sdk-python/compare/v1.5.5...v1.5.6) (2026-11-06)


### Bug Fixes

* cover langchain 0.1.81 and llama-index 4.7.0 in the span-context stopgaps (RIUS-1732) ([#257](https://github.com/glassflow/rius-sdk-python/issues/257)) ([ac2e872](https://github.com/glassflow/rius-sdk-python/commit/ac2e87282f65bce9f3e5ce9c0b502c105dd4e88f))
* cover langchain 0.1.82 and llama-index 4.7.2 in the span-context stopgaps ([#268](https://github.com/glassflow/rius-sdk-python/issues/268)) ([fa036c1](https://github.com/glassflow/rius-sdk-python/commit/fa036c1d79b3569fdf6a5f58ac6fa98cb9cadffb))

## [1.5.5](https://github.com/glassflow/rius-sdk-python/compare/v1.5.4...v1.5.5) (2026-10-27)


Expand Down
16 changes: 14 additions & 2 deletions README.md
Original file line number Diff line number Diff line change
Expand Up @@ -94,7 +94,7 @@ variables:
| `service_version` | `RIUS_SERVICE_VERSION` | none | Sets `service.version` on the resource. Unset leaves it off, so `OTEL_RESOURCE_ATTRIBUTES` can supply it. |
| `disabled` | `RIUS_DISABLED` | `false` | Kill switch. When true, spans are created but never exported. |
| `sample_rate` | `RIUS_SAMPLE_RATE` | `1.0` | Head sampling ratio `0.0`–`1.0` (whole-trace; children follow root). |
| `capture_content` | `RIUS_CAPTURE_CONTENT` | `true` | When false, prompt/response content is stripped at export (metadata still sent). |
| `capture_content` | `RIUS_CAPTURE_CONTENT` | `true` | When false, prompt/response content is stripped at export (metadata still sent). When `RIUS_CAPTURE_CONTENT` is unset, `OTEL_INSTRUMENTATION_GENAI_CAPTURE_CONTENT` is read (see below). |
| `bridge_foreign_provider` | `RIUS_BRIDGE_FOREIGN_PROVIDER` | `false` | Also export the spans of another OpenTelemetry SDK that owns the global provider (see below). |
| `agent_name` | `RIUS_AGENT_NAME` | `service_name` | The agent this process is. Also the `gen_ai.main_agent.id` when `main_agent_id` is unset, so renaming it changes that id. |
| `main_agent_id` | `RIUS_MAIN_AGENT_ID` | a `gen_ai.main_agent.id` in `OTEL_RESOURCE_ATTRIBUTES`, else the agent name | A stable id for the agent (a registry id or an ARN), stamped as `gen_ai.main_agent.id`. Never a per-process value. |
Expand Down Expand Up @@ -149,6 +149,18 @@ rius.init(mask=lambda value, *, key: hash_pii(value) if "input" in key else valu
rius.init(capture_content=False) # drop content entirely, keep metadata
```

Content capture also follows the OpenTelemetry GenAI setting
`OTEL_INSTRUMENTATION_GENAI_CAPTURE_CONTENT` when `RIUS_CAPTURE_CONTENT` is
unset. `NO_CONTENT` turns capture off, and so does `EVENT_ONLY`, because Rius
records content on spans. `SPAN_ONLY` and `SPAN_AND_EVENT` keep it on. The value
is case-insensitive. An unrecognised one keeps capture on and logs a warning
naming the valid values (the OpenTelemetry conventions fall back to
`NO_CONTENT` instead). An empty value of either variable counts as unset (an
empty `RIUS_CAPTURE_CONTENT` also logs a warning).
The `capture_content` argument wins over both variables, and
`RIUS_CAPTURE_CONTENT` wins over the OpenTelemetry one. With neither set,
capture is on, unlike the OpenTelemetry default of `NO_CONTENT`.

### Sessions, users and workspaces

Three context managers scope every span started inside them, including the
Expand Down Expand Up @@ -247,7 +259,7 @@ which it already is in any project that calls Anthropic. When an
instrumentation's library is missing or too old, `init()` logs a warning naming
the version it needs and leaves that one integration off. For example,
`openinference-instrumentation-anthropic` 3.x needs `anthropic>=1.8.0`; with
an older `anthropic`, upgrade it or keep the instrumentation on 2.x.
an older `anthropic`, upgrade it.

```python
rius.init() # auto-enables installed instrumentors
Expand Down
4 changes: 2 additions & 2 deletions pyproject.toml
Original file line number Diff line number Diff line change
Expand Up @@ -57,13 +57,13 @@ dependencies = [
# installed `mcp` would be the one place the SDK added a runtime library to a
# user's environment, which is exactly what the rule above exists to prevent.
openai = ["openinference-instrumentation-openai>=0.1.52"]
anthropic = ["openinference-instrumentation-anthropic>=2.1.0"]
anthropic = ["openinference-instrumentation-anthropic>=3.1.1"]
langchain = ["openinference-instrumentation-langchain>=0.1.67"]
llama-index = ["openinference-instrumentation-llama-index>=4.4.3"]
litellm = ["openinference-instrumentation-litellm>=0.1.34"]
instruments = [
"openinference-instrumentation-openai>=0.1.52",
"openinference-instrumentation-anthropic>=2.1.0",
"openinference-instrumentation-anthropic>=3.1.1",
"openinference-instrumentation-langchain>=0.1.67",
"openinference-instrumentation-llama-index>=4.4.3",
"openinference-instrumentation-litellm>=0.1.34",
Expand Down
Loading
Loading