oyasumi c5fca380ef fix(safety): close three evidence bypasses and make workspace staging idempotent ... Grouped browser verbs were classified by their verb alone. `tab` and `session` sit in the passive set, so `tab new <url>` — documented as navigating — and `session clear` earned a deterministic allow and executed unreviewed in guarded mode and unblocked in observe, while `open <url>`, the same navigation, was reviewed. Passivity is now decided from verb plus subcommand, and the packet carries the result so observe mode stops maintaining a second copy of the rule that could drift more permissive than guarded. The blocked-action list still matches on the bare verb, so `auth login` keeps matching `auth`. Interpreters were a seven-name allowlist, so `python3.12`, `/usr/bin/python3`, `php`, and `deno` set no script path and produced a packet with no artifacts that was still stamped complete — the exact shape the reviewer is told it may allow. Recognize versioned and common interpreters so their sources are actually collected, and fail closed when a command runs code that cannot be resolved to an inspectable script. `from pkg import payload` collected only the package initializer, because an imported name was treated as an attribute and never as a submodule. Effectful code in `pkg/payload.py` executed without appearing in the evidence. Workspace staging runs twice per run and was not idempotent: the second pass read the origin from `source_path`, which the first pass had already rewritten to the copy. With the completion marker absent it cleared the destination and then copied from that same emptied directory, silently handing the agent an empty workspace. The origin is now read back from `original_source_path`. Each fix is covered by a test that fails when the fix is reverted. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>		2026-08-08 00:47:21 +00:00
..
__init__.py	Add configurable token / cost usage limits (#576)	2026-06-22 11:17:08 -04:00
test_agent_factory_shell.py	feat(safety): add contextual action review with guarded and observe modes	2026-08-08 00:47:21 +00:00
test_agent_factory_tool_arguments.py	fix(tools): accept both the string and structured form of every tool argument (#957)	2026-08-01 18:53:24 -07:00
test_agent_tool_registration.py	refactor(tools): split wait_for_message into respond_to_user + wait_for_agents	2026-08-02 02:15:51 +03:00
test_api_spec.py	feat: support API specs and Postman collections as targets (#866)	2026-08-03 21:07:44 -07:00
test_api_spec_targets.py	feat: support API specs and Postman collections as targets (#866)	2026-08-03 21:07:44 -07:00
test_auth_cli.py	Sign in with a ChatGPT subscription for inference (#854)	2026-07-24 15:41:19 -07:00
test_cli_target_list.py	feat(tui): replace Textual with a Go/Bubble Tea interface (#941)	2026-08-03 19:23:07 -07:00
test_codex_auth.py	feat(tui): replace Textual with a Go/Bubble Tea interface (#941)	2026-08-03 19:23:07 -07:00
test_codex_streaming.py	Sign in with a ChatGPT subscription for inference (#854)	2026-07-24 15:41:19 -07:00
test_compaction.py	fix(llm): pass LLM_EXTRA_HEADERS through ModelSettings so they reach the agent loop (#937)	2026-07-29 19:38:06 -07:00
test_config_loader.py	fix(safety): gate browser safety guidance on the active mode, and close test gaps	2026-08-08 00:47:21 +00:00
test_context_budget.py	fix(llm): avoid auth during ChatGPT lookup	2026-07-31 03:45:41 +03:00
test_cost_tracking.py	feat(tui): replace Textual with a Go/Bubble Tea interface (#941)	2026-08-03 19:23:07 -07:00
test_dedupe_model.py	fix(llm): pass LLM_EXTRA_HEADERS through ModelSettings so they reach the agent loop (#937)	2026-07-29 19:38:06 -07:00
test_disable_streaming.py	fix(llm): cap the tool calls one assistant response may queue (#977)	2026-08-06 00:06:59 +03:00
test_docker_client_delete.py	fix(runtime): swallow torn-down docker socket in sandbox delete() (#721)	2026-07-10 00:13:35 -04:00
test_e2e_budget_lifecycle.py	refactor(tools): split wait_for_message into respond_to_user + wait_for_agents	2026-08-02 02:15:51 +03:00
test_execution.py	fix(core): settle a non-interactive agent's status before its exception unwinds	2026-08-04 06:14:54 +03:00
test_execution_transient_retry.py	fix: pre-v1-style lifecycle resilience — mailbox delivery, uniform revival, unexitable runner, waiting timeout, broader retries, crash-safe identity (#923)	2026-08-01 11:17:08 -07:00
test_fenced_code.py	refactor: move strix/viewer under strix/interface	2026-07-26 13:11:14 -07:00
test_go_tui_runtime.py	feat(tui): replace Textual with a Go/Bubble Tea interface (#941)	2026-08-03 19:23:07 -07:00
test_hooks.py	feat(runtime): graduated wrap-up warnings, budget reserve, and interactive budget pause/continue (#893)	2026-07-26 20:37:14 -07:00
test_inputs.py	feat(tui): replace Textual with a Go/Bubble Tea interface (#941)	2026-08-03 19:23:07 -07:00
test_install_script.py	Add Linux ARM64 standalone release support (#886)	2026-07-26 16:27:02 -07:00
test_list_reports.py	feat(reporting): add read-only list_reports + get_report tools (#889)	2026-07-26 14:05:53 -07:00
test_llm_extra_headers.py	fix(llm): apply LLM_EXTRA_HEADERS on native OpenAI route even without a custom base	2026-07-30 04:13:25 +03:00
test_local_sources.py	feat(tui): replace Textual with a Go/Bubble Tea interface (#941)	2026-08-03 19:23:07 -07:00
test_model_retry.py	Sign in with a ChatGPT subscription for inference (#854)	2026-07-24 15:41:19 -07:00
test_models.py	Sign in with a ChatGPT subscription for inference (#854)	2026-07-24 15:41:19 -07:00
test_notes.py	feat(reporting): add read-only list_reports + get_report tools (#889)	2026-07-26 14:05:53 -07:00
test_optional_deps.py	fix(providers): declare bedrock + vertex extras and add provider import-error hints (#588)	2026-07-07 10:24:49 -04:00
test_output_store.py	feat(context): spill oversized tool output into the sandbox workspace (#882)	2026-07-26 14:42:22 -07:00
test_packaging.py	feat(tui): replace Textual with a Go/Bubble Tea interface (#941)	2026-08-03 19:23:07 -07:00
test_provider_hints.py	fix(providers): match google submodule imports and walk full exception chain	2026-07-10 07:21:47 -07:00
test_proxy_client.py	strip transfer encoding (#820)	2026-07-20 21:56:34 -04:00
test_report_pdf.py	refactor: move strix/viewer under strix/interface	2026-07-26 13:11:14 -07:00
test_report_writer.py	fix(reports): safe-fence markdown PoC export and share fence helpers	2026-07-25 13:09:57 -07:00
test_reporting_fields.py	feat(report): keep dependency findings from distinct manifests separate in dedupe	2026-08-06 02:20:46 +03:00
test_reporting_tool.py	fix: calibrate vulnerability severity to demonstrated impact	2026-08-03 23:40:32 +03:00
test_respond_to_user.py	refactor(tools): split wait_for_message into respond_to_user + wait_for_agents	2026-08-02 02:15:51 +03:00
test_runner_rate_limit.py	fix(core): recover from hallucinated tool names instead of ending the scan	2026-08-04 06:14:54 +03:00
test_runner_root_prompt.py	fix(core): recover from hallucinated tool names instead of ending the scan	2026-08-04 06:14:54 +03:00
test_safety_evidence.py	fix(safety): close three evidence bypasses and make workspace staging idempotent	2026-08-08 00:47:21 +00:00
test_safety_prompt.py	fix(safety): gate browser safety guidance on the active mode, and close test gaps	2026-08-08 00:47:21 +00:00
test_safety_reviewer.py	fix(safety): gate browser safety guidance on the active mode, and close test gaps	2026-08-08 00:47:21 +00:00
test_safety_runtime.py	fix(safety): close three evidence bypasses and make workspace staging idempotent	2026-08-08 00:47:21 +00:00
test_sarif.py	feat(report): SARIF 2.1.0 emitter for CI / code-scanning integration (#626)	2026-07-03 10:43:31 -04:00
test_sarif_stride.py	feat(report): tag SARIF rules with STRIDE legs derived from CWE (#708)	2026-07-06 21:19:53 -04:00
test_secret_files.py	Create credential files with owner-only permissions (#945)	2026-08-03 19:36:59 -07:00
test_session_entries.py	feat(runtime): mount local targets instead of copying them in (#958)	2026-08-02 07:45:10 -07:00
test_skill_dir_extension.py	expand firebase storage rules coverage (#1002)	2026-08-06 23:22:06 -07:00
test_state_repo_context.py	fix(report): omit SARIF provenance for multiple repos (#726)	2026-07-10 09:41:18 -04:00
test_stream_idle_timeout.py	fix(llm): abandon a model stream that stops producing events (#978)	2026-08-06 00:07:14 +03:00
test_tool_call_ids.py	fix(llm): cap the tool calls one assistant response may queue (#977)	2026-08-06 00:06:59 +03:00
test_tool_call_limits.py	fix(llm): cap the tool calls one assistant response may queue (#977)	2026-08-06 00:06:59 +03:00
test_tui_backend_controller.py	feat(tui): replace Textual with a Go/Bubble Tea interface (#941)	2026-08-03 19:23:07 -07:00
test_tui_backend_server.py	feat(tui): replace Textual with a Go/Bubble Tea interface (#941)	2026-08-03 19:23:07 -07:00
test_tui_protocol_conformance.py	feat(tui): replace Textual with a Go/Bubble Tea interface (#941)	2026-08-03 19:23:07 -07:00
test_tui_resume_history.py	feat(tui): replace Textual with a Go/Bubble Tea interface (#941)	2026-08-03 19:23:07 -07:00
test_unknown_tool_recovery.py	fix(core): recover from hallucinated tool names instead of ending the scan	2026-08-04 06:14:54 +03:00
test_unraisable_filter.py	feat(tui): replace Textual with a Go/Bubble Tea interface (#941)	2026-08-03 19:23:07 -07:00
test_update_check.py	Add Linux ARM64 standalone release support (#886)	2026-07-26 16:27:02 -07:00
test_usage_subscription.py	Sign in with a ChatGPT subscription for inference (#854)	2026-07-24 15:41:19 -07:00
test_viewer.py	Scope viewer session cookie to the bound port (#922)	2026-07-27 20:37:54 -04:00
test_viewer_auth.py	refactor: move strix/viewer under strix/interface	2026-07-26 13:11:14 -07:00
test_viewer_runs_gating.py	refactor: move strix/viewer under strix/interface	2026-07-26 13:11:14 -07:00
test_wait_dedupe.py	fix(agents): collapse repeated waits queued inside one model turn (#979)	2026-08-06 00:07:30 +03:00
test_workspace_isolation.py	fix(safety): close three evidence bypasses and make workspace staging idempotent	2026-08-08 00:47:21 +00:00