23 Commits

Author SHA1 Message Date
NCBM 5c13d10989 tools: model batch steps as nested executions 2026-11-02 14:21:26 +08:00
NCBM c1b40a7bf4 test: cover options read with getopt
`--` ends a wrapper's options, jammed values (`-p2222`) and launcher flags
(`uv run --no-project pytest`) resolve, and an unknown option stops the unwrap
with the wrapper as the command — where the review asks and the denylist still
sees `sudo`.
2026-10-23 10:47:45 +08:00
NCBM 3cb4d490ed core/tools: announce a remote shell instead of judging it
`ssh prod` runs no command this scan can resolve, and what is typed or piped
into the shell it opens never reaches any gate — `write_pty` and
`run_argv(stdin=…)` are both unchecked input. The tool description said nothing
about it, so the call looked like any other.

The hop is now reported even when no command follows
(`RemoteTarget.shell`), and the review announces it: `remote shell on 'prod'
(ssh with no command; its input is not checked)`. It is a soft confirm like the
other reviews (YOLO may skip it, an escalation behind a hop still may not) — the
client itself stays outside the denylist, since there is no program to name.
2026-10-23 10:43:05 +08:00
NCBM 8c289db545 test: cover the far end of ssh
What the scan resolves (destination, options, jammed `-p2222`, mosh's `--`, a
hop behind `sudo`), what it leaves alone (a client with no command), and both
gates: `check_command_allowed` now refuses `ssh prod rm -rf /` and prompts for
the `sudo` behind the hop, while harmless remote commands stay allowed; the
review prints a remote `bash -c` code block and always asks under
`BatchMode=yes`.
2026-10-23 10:37:14 +08:00
NCBM 2369592473 core/tools+cli/display: name the trash in the delete confirm
The box asked "delete path 'x'", which reads as final even now that it is not.
It now says "move path 'x' to the system trash" — and "delete path 'x'
permanently" when the path really is destroyed: an explicit `permanent=true`, a
host that turned `[trash] enabled` off, or a scratch path under the OS temporary
directory. The classifier resolves the path to answer that, and falls back to
the trash wording when path policy will refuse the call anyway.

The pretty line follows: a delete that landed in the trash prints `(to trash)`
instead of `(done)`, so a session's transcript still shows which deletes could
be undone.
2026-10-21 20:23:54 +08:00
NCBM 8ec63aa6a2 core/agent: let YOLO mode skip YOLO-tagged soft confirms
`--yes` / `/yolo on|off|once` clears the prompt of a `YOLO`-tagged tool: interpreter runs (`run_argv`, `run_argv_batch`, `open_pty`), the file mutations (overwrite / copy / move / delete) and `fetch_local`'s cleartext and mutating requests. A tool without the tag still asks, and hard path/command denylists stay untouched.

`classify_danger` now answers with a `DangerReason` — the reason text plus `yolo_skip` — so one call can refuse that skip: a command that runs through `sudo` / `su` / `doas` / `pkexec` (as the program or a wrapper) keeps its confirm whatever YOLO mode says. It never adds a reason of its own, so the program the denylist gate already asks about is not asked twice.

`prepare_batch_confirm` leaves the covered calls out, so a skipped call neither prompts on its own nor takes a slot in the round's aggregated prompt, and one escalating step keeps the whole batch in front of the human.
2026-10-21 19:10:46 +08:00
NCBM ac33e92d33 core/tools+config: refuse write_file overwrite unless overwrite=true 2026-10-16 18:44:14 +08:00
NCBM 135c28a622 core/tools: review a wrapper option that carries a command line
`env -S "rm -rf /"` ran a command nothing looked at: the option belongs to the
wrapper, so the scan skipped it and the denylist found no program to block while
the soft confirm had no interpreter to review. Collect those values — `env
-S`/`--split-string` and `su -c`/`--command`/`--session-command`, jammed
`--opt=value` and `-Svalue` forms included — and confirm the call the way an
interpreter's `-c` one-liner is confirmed, printing the text under its own
`env -S` label. Recursing into the text for a hard denylist is left out for now:
the review asks the human, who sees the command.
2026-10-06 21:54:18 +08:00
NCBM 2b3b2073a0 core/tools: only value-taking options may swallow the program
The wrapper option table listed flags as if they consumed the next token, so
the program behind them was skipped: `exec -c git -c init.defaultBranch=main
init` resolved to `init` and the policy denied it as a denylisted basename,
while `exec -c rm -rf /`, `command -p rm -rf /` and `command -v rm` hid `rm`
from the denylist completely and `exec -c python -c …` lost the interpreter
review. Drop the entries that take no value — `exec -c`/`-l`, the whole
`command` entry, `pkexec --disable-internal-agent`, `su -p`, and `xargs -e`
(its argument is optional, and the jammed form needs no entry) — so the
resolved program is the one that really runs.
2026-10-06 21:50:21 +08:00
NCBM 7862a555bc core/tools: review interpreters and detached runners past wrappers 2026-10-06 19:44:29 +08:00
NCBM df8f6bf959 core/tools: rename run_command_batch to run_argv_batch (steps/argv keys) 2026-08-04 21:40:25 +08:00
NCBM d698fbf06f core/tools: rename run_command to run_argv (argv param) 2026-08-04 21:38:20 +08:00
NCBM 1d718582ac core/tools: drop process workspace/todo globals; require bound state 2026-07-24 16:32:33 +08:00
NCBM 12c8adda3a test/tools: catalog, grants, and async tool helpers 2026-07-24 14:57:03 +08:00
NCBM 7bdd9ff8b7 chore: pin ruff and basedpyright for prek and lockfile
Match prek additional_dependencies and pyproject dev pins to pdm.lock
(ruff==0.15.22, basedpyright==1.39.9). Format fix for ruff 0.15.22.
2026-07-20 04:35:26 +08:00
NCBM a54fa534f3 test/tools: clean shell confirm format test 2026-07-19 13:05:43 +08:00
NCBM fa9f28687f test/tools: fix shell confirm format test syntax 2026-07-19 13:02:24 +08:00
NCBM 95f277eda7 core/tools: shell confirm argv uses $(command); full -c body indented 2026-07-19 13:01:06 +08:00
NCBM f13ea9f316 core/tools: soft-confirm file overwrite only, not partial edits 2026-07-19 03:05:01 +08:00
NCBM f51d11d46f core/tools/process: add run_command_batch with pipe_out and mix_stderr 2026-07-19 02:46:21 +08:00
NCBM b1a8c23582 core/tools: soft-confirm shells, REPLs, and python/bash -c one-liners
classify_danger prompts on interactive shell/runtime launches and any -c
code (with argv + snippet preview). Confirm deny can attach a user comment
for the model. Restores ToolRegistry.execute after WIP breakage.
2026-07-18 17:41:44 +08:00
NCBM aaf7fe8e7b ci/lint: ruff format pass and fix compact seed test typing
Apply pending ruff format; narrow AssistantChatMessage.content before
membership checks so basedpyright accepts Unset|None unions.
2026-07-15 11:36:01 +08:00
NCBM 510dd67a54 core/tools: classify destructive tools and clarify policy denies
Add classify_danger for delete/move/overwrite paths, get_path_denylist,
and clearer path/command denial messages.
2026-07-14 23:30:02 +08:00