[AAASM-5798] ✨ (scripts): Roll out the claim-vocabulary gate to python-sdk

.github/workflows/claim-vocabulary-check.yml

@@ -0,0 +1,54 @@
+name: Claim Vocabulary Check
+
+# AAASM-5798 / AAASM-5599: ADR 0033's banned-absolute and undifferentiated-verb
+# rules against this repo's reader-facing Markdown (README, CONTRIBUTING,
+# SECURITY, docs/**). Ported from agent-assembly's check_claim_vocabulary.py
+# (AAASM-5679) and docs Hub's port (AAASM-5797); per AAASM-5795's
+# federation-by-convention decision this repo's copy is independent -- see
+# scripts/check_claim_vocabulary.py's own docstring.
+#
+# Runs where CLAUDE.md's own "Docs-only PRs run NO CI" gap otherwise leaves
+# docs/**/*.md and README.md changes with nothing but human review: this repo's
+# main ci.yaml path filter excludes docs/** and *.md entirely.
+#
+# Full-tree, not diff-scoped: the tree's blocking baseline is zero as of this
+# same PR (five real findings fixed alongside the port), so per
+# claim-vocabulary.md's adoption sequence there is no backlog to work down
+# with --diff-base.
+
+on:
+  pull_request:
+    paths:
+      - "README.md"
+      - "CONTRIBUTING.md"
+      - "SECURITY.md"
+      - "docs/**"
+      - "scripts/check_claim_vocabulary.py"
+      - ".github/workflows/claim-vocabulary-check.yml"
+  push:
+    branches:
+      - main
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
+permissions:
+  contents: read
+
+jobs:
+  claim-vocabulary:
+    name: Check for banned-absolute and undifferentiated-verb claims
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7
+
+      - uses: actions/setup-python@5fda3b95a4ea91299a34e894583c3862153e4b97 # v7.0.0
+        with:
+          python-version: "3.12"
+
+      - name: Run the claim-vocabulary self-test
+        run: python3 scripts/check_claim_vocabulary.py --selftest
+
+      - name: Scan for banned-absolute and undifferentiated-verb claims
+        run: python3 scripts/check_claim_vocabulary.py --root .

README.md

@@ -177,7 +177,7 @@ with init_assembly(
 
 What this does:
 
-1. `init_assembly()` registers the agent with the gateway and auto-loads the LangChain adapter — every tool call from now on goes through the policy gate.
+1. `init_assembly()` registers the agent with the gateway and auto-loads the LangChain adapter — each tool call the wrapped agent makes from now on goes through the policy gate.
 2. The `FakeListLLM` replays canned responses so the example runs **offline** with no real LLM.
 3. The `with` block tears down the gateway connection and unwinds adapter hooks on exit.
 

docs/concepts/index.md

@@ -127,7 +127,7 @@ Two independent knobs control governance. It's worth keeping them straight:
 | --- | --- |
 | `auto` (default) | Picks the best available layer for the current platform (eBPF on Linux, else proxy). |
 | `sdk-only` | In-process only — framework adapters enforce on tool calls; no network sidecar. Most portable; best for tests. |
-| `proxy` | Routes outbound traffic through the `aasm` sidecar proxy — network-egress policy with no code changes. |
+| `proxy` | Routes outbound traffic through the `aasm` sidecar proxy — network-egress policy without modifying the agent's source. |
 | `ebpf` | Kernel-level interception via eBPF. **Linux only** — raises `ConfigurationError` elsewhere. |
 
 ### Enforcement modes

docs/examples/langchain-basic-agent.md

@@ -4,7 +4,7 @@ Integrate Agent Assembly with LangChain to enforce governance policy on tool cal
 
 ## What this example demonstrates
 
-This example wires Agent Assembly governance into a LangChain agent so that every tool call is checked against policy *before* the tool runs. It covers:
+This example wires Agent Assembly governance into a LangChain agent so that each tool call the agent makes is checked against policy *before* the tool runs. It covers:
 
 - Initializing Agent Assembly with `init_assembly()`.
 - Wrapping LangChain tools with `AssemblyCallbackHandler` + a governance interceptor.

docs/examples/llamaindex-tool-policy.md

@@ -1,6 +1,6 @@
 # LlamaIndex
 
-Integrates Agent Assembly with [LlamaIndex](https://docs.llamaindex.ai/) using the **native `LlamaIndexAdapter`**, so every tool call a LlamaIndex agent makes is governed automatically — no per-tool wrapper.
+Integrates Agent Assembly with [LlamaIndex](https://docs.llamaindex.ai/) using the **native `LlamaIndexAdapter`**, so each tool call a LlamaIndex agent makes is governed automatically — no per-tool wrapper.
 
 ## What this example demonstrates
 

docs/index.md

@@ -1,7 +1,7 @@
 # Agent Assembly Python SDK
 
 **In plain terms:** this SDK is how a Python agent asks for permission before it acts.
-You wrap your existing agent in one `init_assembly()` call, and from that point on every
+You wrap your existing agent in one `init_assembly()` call, and from that point on each
 tool call your agent makes is checked against a governance policy — allowed or denied —
 without you rewriting a single line of the agent itself. Over a connected runtime the
 outcome of each call is also handed to the runtime's event channel (see below).