jason.woltjeandClaude Opus 5.5
d6ea568312
Ledger skips T3 reasoning summaries and reads issue pages with a
10-page ceiling. Candidate manifest, patch against cc48b124f, mutants
(both killed) and a passing full gate.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
89 lines
7.0 KiB
Plaintext
89 lines
7.0 KiB
Plaintext
✔ sessionModel: agent vars win, then the system's execution settings (13.592456ms)
|
||
✔ a pi bundle: prompt, policy, tools and manifest, 0600 in a 0700 directory (6.300387ms)
|
||
✔ a claude-code bundle adds the wrapped gate hook and the MCP config (3.429935ms)
|
||
✔ the prompt's built-in tools line is the harness's own names, as the manifest lists them (9.261754ms)
|
||
✔ the prompt names the project clones in the workspace, and none when there are none (5.397023ms)
|
||
✔ the prompt explains the review loop to a role with push_working, and reviews to one with fetch_review (5.832209ms)
|
||
✔ a bundle is written once: an existing file refuses (1.351598ms)
|
||
✔ a path with a single quote can't go into the hook command (1.389334ms)
|
||
✔ allow exits 0, a deny exits 2 with the reason on stderr (131.919529ms)
|
||
✔ a missing or wrong policy, or a bad event, exits 2 (74.22573ms)
|
||
✔ the bundle's wrapped command: a missing gate or node still blocks (1088.044553ms)
|
||
✔ claude: typed tools through MCP, the hook blocks, builtins outside --tools don't exist (734.592609ms)
|
||
✔ claude: the hook alone blocks a path outside the workspace (443.02577ms)
|
||
✔ claude: a second turn resumes the first turn's session (740.873375ms)
|
||
✔ claude adapter: --restricted is always passed (5.275618ms)
|
||
✔ claude adapter: MOSAIC_CLAUDE_BIN wins over PATH and must be an absolute executable (20.496068ms)
|
||
✔ claude adapter: a directory pin exits 2 before it runs; a pin with a space runs (6.059446ms)
|
||
✔ claude: CLAUDE.md files and auto-memory don't reach the model; without --restricted they do (727.691785ms)
|
||
✔ claude: a missing hook or MCP file refuses before claude starts (7.347032ms)
|
||
✔ pi: policy tools and typed tools pass, anything else is blocked (3.141994ms)
|
||
✔ claude: builtins map from pi names, typed tools need the mcp prefix (0.716647ms)
|
||
✔ file tool paths must resolve inside the workspace (1.324356ms)
|
||
✔ pi's own path normalisation can't be used to step out (1.024039ms)
|
||
✔ a symlink inside the workspace that points out is outside (0.799356ms)
|
||
✔ a dangling symlink is refused at any depth, in both harnesses (2.613665ms)
|
||
✔ read is checked under every spelling pi's read would open, in both harnesses (15.982849ms)
|
||
✔ other spellings cover directories, dangling links and pi's cwd (1.463989ms)
|
||
✔ a relative path climbs from the workspace's real path, in both harnesses (4.821926ms)
|
||
✔ claude path fields per tool (0.73711ms)
|
||
✔ glob patterns stay inside the workspace (0.650598ms)
|
||
✔ a path that can't be checked is blocked (0.360691ms)
|
||
✔ fetch_review's have lists the working refs before origin's, and no more than HAVE_MAX (211.888205ms)
|
||
✔ fetch_review's have orders the working refs newest first within each clone, one clone after another (140.135958ms)
|
||
✔ a runner git past its time: SIGTERM reaches its process group, and SIGKILL follows after the grace period (1809.722696ms)
|
||
✔ once a signalled runner git has closed, its group gets SIGKILL, so a process that ignored SIGTERM dies too (303.967694ms)
|
||
✔ initialize, ping and tools/list (45.825924ms)
|
||
✔ tools/call goes through the tool socket; a refusal is an isError result (36.070466ms)
|
||
✔ replies are compact JSON; over the cap a tool error names the size and the cap (38.747852ms)
|
||
✔ unknown tools and methods are JSON-RPC errors and never reach the socket (28.653523ms)
|
||
✔ a missing argument is a usage error (23.129387ms)
|
||
✔ pi: typed tools reach the socket, the gate blocks, agent_end writes the marker (366.921294ms)
|
||
✔ pi: a reply over the cap is a tool error naming the size and the cap; list_tasks asks for open tasks (316.115261ms)
|
||
✔ pi: a write through a dangling symlink is blocked, and nothing appears outside (319.328383ms)
|
||
✔ pi: a read is refused when pi would open another spelling outside (312.088811ms)
|
||
✔ pi: a relative path climbs from the real path of a workspace behind a symlink (304.802288ms)
|
||
✔ pi: a relative path climbs from the real path of a dataRoot behind a symlink (330.080833ms)
|
||
✔ pi: a missing extension refuses before any model call (5.812296ms)
|
||
✔ pi: an extension without its configuration fails pi's start (247.300778ms)
|
||
✔ founderCheck: founder variables, then a needed service without a usable token (1.708404ms)
|
||
✔ turnRequest names the sender, class, reply and decision (0.304737ms)
|
||
✔ turnRequest says a tester's message is a test and never Jason's instruction (0.306605ms)
|
||
✔ turnRequest shows the message's own sent time, a day old as well as new (0.21075ms)
|
||
✔ stateBlock names the launch state with its time and event, and the open tasks with the cap (0.313829ms)
|
||
✔ a message becomes a turn, the answer goes back as a RESULT, SIGTERM releases and exits 0 (249.280014ms)
|
||
✔ launches revoked and then restored: the turn's block reads allowed and gives the restore time (258.959367ms)
|
||
✔ a task closed on the tracker is missing from the next turn's open list (207.427361ms)
|
||
✔ a message read a day after it was sent shows the time it was sent (143.587533ms)
|
||
✔ a failed state read fails the turn: no adapter runs, the sender hears why, the message is read (118.07172ms)
|
||
✔ a SIGTERM before the claim stops the runner with exit 0 and no claim (101.202114ms)
|
||
✔ typed tools carry the runner's capability; launch goes to the host's launch socket (374.570724ms)
|
||
✔ a RESULT gets no automatic reply; failed turns reply with the reason (1302.463466ms)
|
||
✔ SIGTERM during a turn kills the turn's process group and still exits 0 (145.963685ms)
|
||
✔ founder credentials stop before the claim (20) (161.099363ms)
|
||
✔ a refused claim exits 21; an ended run's capability exits 22 (185.188186ms)
|
||
✔ the launch ending under a running session exits 22 (146.094775ms)
|
||
✔ a broker that stays unreachable exits 23 after brokerRetries polls (242.190245ms)
|
||
✔ a broker that is down at the claim exits 23, not 21 (96.302758ms)
|
||
✔ no capability, or a malformed one, on stdin exits 2 (196.055577ms)
|
||
✔ a missing or malformed policy exits 2 before the claim (140.483167ms)
|
||
✔ the PM gets launch, its task verbs and the reads (7.387137ms)
|
||
✔ list_requirements reads the requirement ids, for a task.create holder with a tracker (4.005451ms)
|
||
✔ a coder gets no launch, no resolve_decision, and no task tools without a tracker (2.110678ms)
|
||
✔ a coder with a tracker gets its own task tools and no task_reassign (1.364625ms)
|
||
✔ launch only when the business's launch block names the instance as launcher (1.91319ms)
|
||
✔ an action outside the instance's authority has no tool (1.752901ms)
|
||
✔ the review loop: push_working and request_review for the coder, fetch_review and review_verdict for the reviewer, none for the PM (4.923006ms)
|
||
✔ callTool: one JSON line out, the result back, a refusal rejects (7.735293ms)
|
||
✔ list_tasks takes the broker's filters and pages; get_task takes a task_ref (2.905418ms)
|
||
✔ callTool sends list_tasks done false unless the session names done; other tools as given (2.636329ms)
|
||
✔ replyText: compact JSON up to the cap, an error naming the size and the cap over it (0.797487ms)
|
||
ℹ tests 80
|
||
ℹ suites 0
|
||
ℹ pass 80
|
||
ℹ fail 0
|
||
ℹ cancelled 0
|
||
ℹ skipped 0
|
||
ℹ todo 0
|
||
ℹ duration_ms 11024.709757
|