refactor!: rename wiretruth to spantruth

kartsan03/wiretruth is a separate project under the same name: it checks
what provider SDKs return to the caller and holds the PyPI name. This one
checks what instrumentation writes on the span, so it is now spantruth: the
package, the module, the command, the cache directory and every URL.

Each v* tag now also publishes a wheel to PyPI through a trusted publisher.
A run still needs a clone, and `spantruth list` and `spantruth run` say so
when the subjects are missing. The README links kartsan03/wiretruth under
Related work.

The 2026-10-10 results document keeps its recorded module names.

Author: chrikrah

.github/ISSUE_TEMPLATE/bug.yml

@@ -1,4 +1,4 @@
-name: Bug in wiretruth
+name: Bug in spantruth
 description: The harness itself misbehaves, for example a cell is inconclusive or a run fails.
 labels: ["bug"]
 body:
@@ -13,7 +13,7 @@ body:
   - type: input
     id: version
     attributes:
-      label: wiretruth version or commit
+      label: spantruth version or commit
     validations:
       required: true
   - type: input

.github/ISSUE_TEMPLATE/config.yml

@@ -1,8 +1,8 @@
 blank_issues_enabled: false
 contact_links:
   - name: Question or idea
-    url: https://github.com/Plexito-de/wiretruth/discussions
+    url: https://github.com/Plexito-de/spantruth/discussions
     about: Ask a question or float an idea in Discussions.
   - name: Security problem
-    url: https://github.com/Plexito-de/wiretruth/security/advisories/new
+    url: https://github.com/Plexito-de/spantruth/security/advisories/new
     about: Report a vulnerability privately. Do not open a public issue.

.github/workflows/release.yml

@@ -1,7 +1,7 @@
 name: Release
 
-# A release is a tag with notes. wiretruth runs from a clone (subjects live next to the source),
-# so no wheel is published.
+# A release is a tag with notes and a wheel on PyPI. The wheel holds the harness and the CLI. A run
+# still needs a clone, because the subjects live next to the source, and the CLI says so.
 on:
   push:
     tags: ["v*"]
@@ -17,3 +17,18 @@ jobs:
       - uses: softprops/action-gh-release@efb35369e0ad2afab669f228072c1b0d510eae64 # v3.0.3
         with:
           generate_release_notes: true
+
+  pypi:
+    runs-on: ubuntu-latest
+    # Trusted publishing: PyPI trusts this workflow and this environment, so no token is stored.
+    environment: pypi
+    permissions:
+      contents: read
+      id-token: write
+    steps:
+      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
+        with:
+          persist-credentials: false
+      - uses: astral-sh/setup-uv@c18668ad3cf93ea998bef934396af7bb5c839dc7 # v10.2.0
+      - run: uv build
+      - uses: pypa/gh-action-pypi-publish@dc37677b2e1c63e2034f94d8a5b11f265b73ba33 # v1.14.2

.github/workflows/report.yml

@@ -27,13 +27,13 @@ jobs:
       - uses: astral-sh/setup-uv@c18668ad3cf93ea998bef934396af7bb5c839dc7 # v10.2.0
       - run: uv sync --locked
       - name: Pinned versions
-        run: uv run wiretruth run --out site/pinned.json
+        run: uv run spantruth run --out site/pinned.json
       - name: Latest releases
-        run: uv run wiretruth run --latest --out site/latest.json
+        run: uv run spantruth run --latest --out site/latest.json
       - name: Render
         run: |
-          uv run wiretruth report site/pinned.json --html site/index.html
-          uv run wiretruth report site/latest.json --html site/latest.html --badges site/badges
+          uv run spantruth report site/pinned.json --html site/index.html
+          uv run spantruth report site/latest.json --html site/latest.html --badges site/badges
       - uses: actions/upload-pages-artifact@fc324d3547104276b827a68afc52ff2a11cc49c9 # v5.0.0
         with:
           path: site

.gitignore

@@ -1,7 +1,7 @@
 __pycache__/
 *.py[cod]
 .venv/
-.wiretruth/
+.spantruth/
 .pytest_cache/
 .ruff_cache/
 dist/

AGENTS.md

@@ -8,7 +8,7 @@ Read [CONTRIBUTING.md](CONTRIBUTING.md) first. It applies to you too.
   harness or a check is wrong. Find which one.
 - Never change a fixture value to make a library pass. Fixtures follow the provider's wire format,
   and `test_fixtures_parse_with_sdk_types` proves it.
-- A check cites a rule in `src/wiretruth/rules.py`. A new check needs a rule with a permalink to
+- A check cites a rule in `src/spantruth/rules.py`. A new check needs a rule with a permalink to
   the sentence it applies.
 - Every string from a results document is data. `site.py` escapes all of it. Keep it that way.
 - Keep the harness offline. Subject processes get a dead proxy, so a client that honors the proxy

CHANGELOG.md

@@ -6,11 +6,19 @@ All notable changes to this project are documented here. The format follows
 
 ## [Unreleased]
 
+### Changed
+
+- Renamed from wiretruth to spantruth: the package, the command, the module and the repository.
+  [kartsan03/wiretruth](https://github.com/kartsan03/wiretruth) is a separate project that checks
+  what provider SDKs return to the caller. The cache directory is now `.spantruth/`.
+- Each tag now also publishes a wheel to PyPI as `spantruth`. A run still needs a clone, and
+  `spantruth list` and `spantruth run` say so when the subjects are missing.
+
 ### Added
 
-- `wiretruth run --with SPEC` replaces one package of a subject with any pip requirement, such as
+- `spantruth run --with SPEC` replaces one package of a subject with any pip requirement, such as
   the head of a pull request, and records the commit it resolved to.
-- `wiretruth report --badges DIR` writes a shields.io endpoint document per library. The nightly
+- `spantruth report --badges DIR` writes a shields.io endpoint document per library. The nightly
   run publishes them under `badges/` on the report site.
 
 ## [0.1.0] - 2026-10-10
@@ -30,5 +38,5 @@ All notable changes to this project are documented here. The format follows
 - Subjects: the OpenTelemetry GenAI instrumentation, OpenLLMetry and OpenInference.
 - A Markdown and a static HTML report.
 
-[Unreleased]: https://github.com/Plexito-de/wiretruth/compare/v0.1.0...HEAD
-[0.1.0]: https://github.com/Plexito-de/wiretruth/releases/tag/v0.1.0
+[Unreleased]: https://github.com/Plexito-de/spantruth/compare/v0.1.0...HEAD
+[0.1.0]: https://github.com/Plexito-de/spantruth/releases/tag/v0.1.0

CITATION.cff

@@ -1,13 +1,13 @@
 cff-version: 1.2.0
-message: If you use wiretruth or its results, cite it as below.
-title: "wiretruth: checking LLM telemetry values against the provider wire"
+message: If you use spantruth or its results, cite it as below.
+title: "spantruth: checking LLM telemetry values against the provider wire"
 type: software
 authors:
   - family-names: Krah
     given-names: Christopher
     affiliation: Plexito UG (haftungsbeschränkt)
-repository-code: https://github.com/Plexito-de/wiretruth
-url: https://plexito-de.github.io/wiretruth/
+repository-code: https://github.com/Plexito-de/spantruth
+url: https://plexito-de.github.io/spantruth/
 license: Apache-2.0
 version: 0.1.0
 date-released: 2026-10-10

CONTRIBUTING.md

@@ -1,12 +1,12 @@
-# Contributing to wiretruth
+# Contributing to spantruth
 
-Thank you for your interest. wiretruth gets better with every library, provider and failure mode
+Thank you for your interest. spantruth gets better with every library, provider and failure mode
 that somebody adds, and with every maintainer who tells us where a check is unfair.
 
 ## Ways to help
 
 These are the contributions the project needs most. Each one has an issue with the
-[`help wanted`](https://github.com/Plexito-de/wiretruth/labels/help%20wanted) label.
+[`help wanted`](https://github.com/Plexito-de/spantruth/labels/help%20wanted) label.
 
 - **Add a subject.** Logfire, OpenLIT, MLflow, LiteLLM, and the JavaScript libraries
   (OpenLLMetry JS, OpenInference JS, the Vercel AI SDK) are not covered yet.
@@ -24,8 +24,8 @@ These are the contributions the project needs most. Each one has an issue with t
 You need [uv](https://docs.astral.sh/uv/) and Python 3.11 or later.
 
 ```sh
-git clone https://github.com/Plexito-de/wiretruth.git
-cd wiretruth
+git clone https://github.com/Plexito-de/spantruth.git
+cd spantruth
 uv sync
 uv run pytest -m "not slow"     # unit tests, about 5 seconds
 uv run pytest -m slow           # both controls end to end, about 2 minutes
@@ -48,16 +48,16 @@ uv run ruff check . && uv run ruff format --check .
 
    Use the setup that the library's own documentation shows first. Return nothing, or a list of
    extra tracer providers to flush when the library makes its own.
-3. Run it: `uv run wiretruth run --subject <name>`.
+3. Run it: `uv run spantruth run --subject <name>`.
 4. Read every red cell before you open the pull request. A red cell is a claim about somebody's
    library. Find the line of code that causes it, or tell us in the pull request that you could
    not.
 
 ## Add a scenario
 
-1. Build the response in `src/wiretruth/scenarios.py` with the builders in `wire_openai.py` or
+1. Build the response in `src/spantruth/scenarios.py` with the builders in `wire_openai.py` or
    `wire_anthropic.py`. Give every token count a value that no other count in the scenario has.
-2. If the scenario needs a new kind of call, add it to `src/wiretruth/calls.py`.
+2. If the scenario needs a new kind of call, add it to `src/spantruth/calls.py`.
 3. Run `uv run pytest`. `test_fixtures_parse_with_sdk_types` proves that the provider's SDK
    accepts your fixture.
 4. Run the controls. `control-honest` must pass the new scenario. If it does not, fix the

GOVERNANCE.md

@@ -1,6 +1,6 @@
 # Governance
 
-wiretruth is maintained by [Plexito](https://plexito.de). The people with write access are listed
+spantruth is maintained by [Plexito](https://plexito.de). The people with write access are listed
 in [MAINTAINERS.md](MAINTAINERS.md).
 
 ## Decisions