✔ sessionModel: agent vars win, then the system's execution settings (10.347285ms)
✔ a pi bundle: prompt, policy, tools and manifest, 0600 in a 0700 directory (5.555993ms)
✔ a claude-code bundle adds the wrapped gate hook and the MCP config (3.636062ms)
✔ the prompt's built-in tools line is the harness's own names, as the manifest lists them (11.510057ms)
✔ the prompt names the project clones in the workspace, and none when there are none (6.108332ms)
✔ the prompt explains the review loop to a role with push_working, and reviews to one with fetch_review (7.717726ms)
✔ a bundle is written once: an existing file refuses (2.1771ms)
✔ a path with a single quote can't go into the hook command (2.262674ms)
✔ allow exits 0, a deny exits 2 with the reason on stderr (127.716983ms)
✔ a missing or wrong policy, or a bad event, exits 2 (92.400777ms)
✔ the bundle's wrapped command: a missing gate or node still blocks (1087.274407ms)
✔ claude: typed tools through MCP, the hook blocks, builtins outside --tools don't exist (764.139729ms)
✔ claude: the hook alone blocks a path outside the workspace (446.060417ms)
✔ claude: a second turn resumes the first turn's session (754.94832ms)
✔ claude adapter: --restricted is always passed (5.541475ms)
✔ claude adapter: MOSAIC_CLAUDE_BIN wins over PATH and must be an absolute executable (16.154334ms)
✔ claude adapter: a directory pin exits 2 before it runs; a pin with a space runs (5.6766ms)
✔ claude: CLAUDE.md files and auto-memory don't reach the model; without --restricted they do (740.181938ms)
✔ claude: a missing hook or MCP file refuses before claude starts (7.493881ms)
✔ pi: policy tools and typed tools pass, anything else is blocked (3.98485ms)
✔ claude: builtins map from pi names, typed tools need the mcp prefix (0.988674ms)
✔ file tool paths must resolve inside the workspace (1.820751ms)
✔ pi's own path normalisation can't be used to step out (1.394656ms)
✔ a symlink inside the workspace that points out is outside (1.243779ms)
✔ a dangling symlink is refused at any depth, in both harnesses (4.579334ms)
✔ read is checked under every spelling pi's read would open, in both harnesses (15.567935ms)
✔ other spellings cover directories, dangling links and pi's cwd (2.290934ms)
✔ a relative path climbs from the workspace's real path, in both harnesses (5.598412ms)
✔ claude path fields per tool (0.695139ms)
✔ glob patterns stay inside the workspace (0.880898ms)
✔ a path that can't be checked is blocked (0.374525ms)
✔ fetch_review's have lists the working refs before origin's, and no more than HAVE_MAX (227.061423ms)
✔ fetch_review's have orders the working refs newest first within each clone, one clone after another (137.059677ms)
✔ a runner git past its time: SIGTERM reaches its process group, and SIGKILL follows after the grace period (1810.842274ms)
✔ once a signalled runner git has closed, its group gets SIGKILL, so a process that ignored SIGTERM dies too (305.646243ms)
✔ initialize, ping and tools/list (50.696205ms)
✔ tools/call goes through the tool socket; a refusal is an isError result (36.39016ms)
✔ replies are compact JSON; over the cap a tool error names the size and the cap (38.03129ms)
✔ unknown tools and methods are JSON-RPC errors and never reach the socket (37.815066ms)
✔ a missing argument is a usage error (33.657201ms)
✔ pi: typed tools reach the socket, the gate blocks, agent_end writes the marker (385.070191ms)
✔ pi: a reply over the cap is a tool error naming the size and the cap; list_tasks asks for open tasks (317.576861ms)
✔ pi: a write through a dangling symlink is blocked, and nothing appears outside (312.118473ms)
✔ pi: a read is refused when pi would open another spelling outside (310.701639ms)
✔ pi: a relative path climbs from the real path of a workspace behind a symlink (306.095706ms)
✔ pi: a relative path climbs from the real path of a dataRoot behind a symlink (317.774581ms)
✔ pi: a missing extension refuses before any model call (5.983977ms)
✔ pi: an extension without its configuration fails pi's start (267.17437ms)
✔ founderCheck: founder variables, then a needed service without a usable token (1.188385ms)
✔ turnRequest names the sender, class, reply and decision (0.186339ms)
✔ turnRequest says a tester's message is a test and never Jason's instruction (0.224418ms)
✔ turnRequest shows the message's own sent time, a day old as well as new (0.140576ms)
✔ stateBlock names the launch state with its time and event, and the open tasks with the cap (0.242242ms)
✔ a message becomes a turn, the answer goes back as a RESULT, SIGTERM releases and exits 0 (265.727605ms)
✔ launches revoked and then restored: the turn's block reads allowed and gives the restore time (235.965959ms)
✔ a task closed on the tracker is missing from the next turn's open list (217.465491ms)
✔ a message read a day after it was sent shows the time it was sent (175.110743ms)
✔ a decision the human resolves becomes a turn citing the decision and the answer (213.488065ms)
✔ a failed state read fails the turn: no adapter runs, the sender hears why, the message is read (118.491432ms)
✔ a SIGTERM before the claim stops the runner with exit 0 and no claim (102.71037ms)
✔ typed tools carry the runner's capability; launch goes to the host's launch socket (422.096172ms)
✔ a RESULT gets no automatic reply; failed turns reply with the reason (1293.239494ms)
✔ SIGTERM during a turn kills the turn's process group and still exits 0 (150.156771ms)
✔ founder credentials stop before the claim (20) (167.45365ms)
✔ a refused claim exits 21; an ended run's capability exits 22 (201.213691ms)
✔ the launch ending under a running session exits 22 (144.457268ms)
✔ a broker that stays unreachable exits 23 after brokerRetries polls (253.963818ms)
✔ a broker that is down at the claim exits 23, not 21 (101.405395ms)
✔ no capability, or a malformed one, on stdin exits 2 (241.453984ms)
✔ a missing or malformed policy exits 2 before the claim (148.41295ms)
✔ the PM gets launch, its task verbs and the reads (8.682596ms)
✔ list_requirements reads the requirement ids, for a task.create holder with a tracker (4.856353ms)
✔ a coder gets no launch, no resolve_decision, and no task tools without a tracker (1.891822ms)
✔ a coder with a tracker gets its own task tools and no task_reassign (1.343874ms)
✔ launch only when the business's launch block names the instance as launcher (1.904065ms)
✔ an action outside the instance's authority has no tool (1.851432ms)
✔ the review loop: push_working and request_review for the coder, fetch_review and review_verdict for the reviewer, none for the PM (6.996691ms)
✔ callTool: one JSON line out, the result back, a refusal rejects (11.272327ms)
✔ list_tasks takes the broker's filters and pages; get_task takes a task_ref (2.55634ms)
✔ callTool sends list_tasks done false unless the session names done; other tools as given (2.468354ms)
✔ replyText: compact JSON up to the cap, an error naming the size and the cap over it (0.673375ms)
ℹ tests 81
ℹ suites 0
ℹ pass 81
ℹ fail 0
ℹ cancelled 0
ℹ skipped 0
ℹ todo 0
ℹ duration_ms 11325.495606