diff --git a/internal/rules/policies_test.go b/internal/rules/policies_test.go index 65af6c0a..756e249c 100644 --- a/internal/rules/policies_test.go +++ b/internal/rules/policies_test.go @@ -1986,6 +1986,21 @@ def fetch_data(x: str) -> dict: return {} `, toolConfig: nil, wantFires: false}, + {name: "AG2-019 fires on a generic tool name", ruleID: "AG2-019", kind: models.KindAutoGenTool, src: ` +def process(payload: str) -> str: + """Handle an incoming payload and return the result of handling it.""" + return payload +`, wantFires: true}, + {name: "AG2-019 silent on a verb-object tool name", ruleID: "AG2-019", kind: models.KindAutoGenTool, src: ` +def summarize_invoice(invoice_id: str) -> str: + """Summarize one invoice by its identifier and return the summary text.""" + return invoice_id +`, wantFires: false}, + {name: "AG2-019 silent on a name merely containing a listed word", ruleID: "AG2-019", kind: models.KindAutoGenTool, src: ` +def process_invoice_batch(batch_id: str) -> str: + """Process one batch of invoices and return a per-invoice result summary.""" + return batch_id +`, wantFires: false}, } // policyRepoRuleCases covers repo-scoped rules. @@ -2246,7 +2261,6 @@ var policyRepoRuleCases = []policyRepoCase{ }, models.RepoInventory{SDKsDetected: []models.SDK{models.SDKOpenAIAgents}}, false}, - } // optionsWithPermissionMode builds a ClaudeAgentOptionsDef whose captured diff --git a/testdata/rules-fixture/autogen/tool_definition.yaml b/testdata/rules-fixture/autogen/tool_definition.yaml index 7c64aca7..36289eaf 100644 --- a/testdata/rules-fixture/autogen/tool_definition.yaml +++ b/testdata/rules-fixture/autogen/tool_definition.yaml @@ -54,3 +54,34 @@ rules: Annotate every parameter with a concrete type (str, int, list[str], a typing.Annotated description, etc.). AutoGen propagates these annotations into the schema the model sees, so the model emits correctly-shaped arguments. + + - id: AG2-019 + title: Ambiguous AutoGen tool name + severity: low + confidence: 0.9 + language: python + applies_to: + - autogen_tool + scope: tool + match: + name_in: + - process + - handle + - run + - do + - execute + - perform + - work + - go + - thing + - stuff + explanation: > + The tool name is a generic verb that says nothing about what the tool + acts on. The name sits directly beside the description in what the model + sees, so it is half the selection signal, and a name like process or + handle spends that half on nothing — the model either calls the tool for + the wrong job or passes over it entirely. Each wrong pick costs a full round of the conversation — propose, execute, reply — so an ambiguous name spends the AG2-004 max_round budget rather than the chat reaching an answer. + fix: > + Rename to a verb-object form that names the thing acted on: + summarize_invoice, refund_charge, fetch_order_status. Keep the + description for the detail and let the name carry the subject.