Allow intentional input clears to remain in exported traces instead of dropping every non-redacted blank fill.
Co-authored-by: Cursor <cursoragent@cursor.com>
Tool cancellation is now cooperative end-to-end: the queue stays busy until compensation completes, cancellation is checked cooperatively across the pipeline, the fallback window is compensated too, and cleanup_failed is surfaced explicitly. Includes a gated delayed-compensation E2E test.
Add a per-browser "Control hints" switch to the popup that hides the in-page control overlay (status pill + orange glow, along with its input blocker) while the Agent controls a tab. Preference lives in chrome.storage.local (bsk_control_hints_hidden, default: shown) and is applied live in content scripts via storage.onChanged.
Both popup settings rows share a single Switch component; the hint copy lives in an info-icon tooltip.
Refs #62
chrome.tabs.captureVisibleTab reads back the window surface, which fails
outright on some Windows/Chrome combinations (Chromium
FAILURE_REASON_READBACK_FAILED: "Failed to capture tab: image readback
failed"), leaving full-tab screenshots without any working path.
Capture the full tab via CDP Page.captureScreenshot (fromSurface) when
the visible-tab capture rejects; the CDP path goes through the
renderer BeginFrame pipeline instead of the surface readback. When both
paths fail, return cdp_failed with data.reason=screenshot_capture_failed
carrying both underlying messages, and render a CLI hint that points at
the browser-side readback cause.
Refs #71
The per-user pipe name embedded %USERNAME% verbatim
(\\.\pipe\bsk-daemon-{user}-{hash}). When the username contains
apostrophes (e.g. z'z'f'l'g'y), NPFS behaviour diverges between create
and open: the daemon's CreateNamedPipeW reports success and logs that it
is listening, yet every client CreateFileW on the same name fails with
ERROR_FILE_NOT_FOUND / ERROR_PATH_NOT_FOUND, so the CLI can never reach
the daemon (issue #75).
The name is now hash-only (\\.\pipe\bsk-daemon-{hash}); the hash
already covers user + home, so uniqueness and per-user/BSK_HOME
isolation are unchanged, and no raw username characters can reach NPFS.
The pipe name is exchanged exclusively through daemon.json (the CLI
never computes it), so old/new CLI/daemon combinations stay compatible.
Unit tests pin the safe character set for apostrophe/space/Unicode
usernames and the per-(user, home) uniqueness.
Fixes#75
`bsk session start --no-focus` opens the Agent Window in the background
without stealing focus. The flag threads focused=false through
CLI -> daemon IPC -> tool.session_start -> chrome.windows.create; an
omitted `focused` keeps the extension default (true) so older clients
are unaffected.
Reworked onto current main so it composes with #52 window sizing:
AgentWindowApi.create / SessionManager.start (extension), start_session
(CLI) and the daemon's session start now take options objects/structs
({ size, focused }) instead of colliding positional params.
Co-authored-by: alectimison-maker <alec.timison@gmail.com>
The first-run installers (install.sh, install.ps1) downloaded and
extracted the bsk release archive with no integrity check, even though
the in-app `bsk update` path treats a sha256 checksum as mandatory
(update.rs bails with "does not include a sha256 checksum; cannot
safely auto-update"). The release version.json already carries a
per-asset sha256 (render-version-json.mjs renders with --dist), so the
installers can verify the same way.
Both installers now fetch version.json once (pinned and latest), read
`assets[<platform>].sha256`, and verify the downloaded archive's sha256
before extracting:
- install.sh: POSIX sed to parse the field (no jq dependency), and
sha256sum/shasum to compute the digest.
- install.ps1: Invoke-RestMethod + Get-FileHash.
A checksum mismatch is fatal (fail-closed, matching `bsk update`). A
missing manifest/checksum, or the absence of a sha256 tool, only skips
verification with a warning, so the first-run bootstrap still works on
minimal/offline machines.
Validation: `sh -n` / `bash -n`; functional test of the sed extraction
against a realistic version.json and of compute_sha256 against
sha256sum. The installers are not exercised by CI (no shell harness).
Co-authored-by: Claude <noreply@anthropic.com>
The daemon's periodic update check now installs a newer release instead
of only refreshing the hint cache: it reuses the update.rs install path
(download, sha256 verify, extract, replace), skips the round while any
agent session is live, and otherwise restarts itself into the new binary
(spawns a detached replacement that waits for the old pid, then exits).
On by default; BSK_AUTO_UPDATE=off keeps the old cache/hint-only
behaviour. Windows can only stage the replacement, so it just logs that
a restart is needed. The CLI hint wording reflects the switch.
* fix(update): run periodic update check in daemon so hints actually fire
The CLI spawned a detached thread to refresh ~/.bsk/update-check.json,
but the process exits as soon as the command finishes, killing the
thread before the fetch could complete — so the cache was never
written and the new-version hint never showed.
Move the periodic check into the daemon: a tokio task ticks every
30 minutes (down from 24h), refreshes the cache when missing/stale,
and logs failures at warn/info into the daemon log. The CLI path is
now read-only: it prints the hint only from a fresh cache and never
spawns threads or touches the network.
* fix(update): decouple daemon refresh window so checks truly run every 30min
The refresh decision reused the 30min tick interval as its freshness
window, and with is_fresh()'s inclusive bound a cache written just
after tick N still counted as fresh at tick N+1 — so in steady state
the daemon only refetched every other tick (~60min).
Judge staleness against DAEMON_REFRESH_WINDOW (25min = 5/6 of the
tick) instead: every 30min tick now finds the cache stale and really
refreshes, while a daemon restarted with a younger-than-25min cache
still skips its first tick. CLI hint freshness semantics
(UPDATE_CHECK_INTERVAL = 30min) are unchanged.
* feat(install-skill): add Kimi Code harness support
Add kimi-code to the install-skill harness list: installs the bundled SKILL.md into ~/.kimi/skills (Kimi Code CLI's brand-group user skills directory, highest priority per https://github.com/MoonshotAI/kimi-cli/blob/main/docs/en/customization/skills.md). Detection signals: ~/.kimi directory and kimi/kimi-cli on PATH. Aliases: kimi, kimi-code, kimi-cli, kimicode, kimi_cli.
* fix(install-skill): point kimi-code harness at ~/.kimi-code and honor KIMI_CODE_HOME
Kimi Code CLI's canonical repo is MoonshotAI/kimi-code: user-level skills live in $KIMI_CODE_HOME/skills (default ~/.kimi-code/skills), not kimi-cli's ~/.kimi/skills. Detection checks the kimi-code home dir (env-aware label) and the kimi binary on PATH. Serialize env-mutating kimi tests with a shared mutex to avoid a parallel-test race on KIMI_CODE_HOME.
* feat(emulate): add mobile device emulation via CDP Emulation domain
Add `bsk emulate` so agents can debug mobile page behaviour by
emulating a device environment on a session tab:
- Viewport metrics (Emulation.setDeviceMetricsOverride), User-Agent
(Emulation.setUserAgentOverride, optional accept-language and UA
client-hints metadata), and touch (Emulation.setTouchEmulationEnabled
with maxTouchPoints).
- Seven CLI-side device presets (iphone-14, iphone-14-pro-max,
iphone-se, pixel-7, galaxy-s23, ipad-mini, galaxy-tab-s8) via
--device; manual flags (--width/--height/--dpr/--mobile/--ua/
--accept-language/--touch/--max-touch-points) work standalone or
override individual preset fields.
- `bsk emulate --off` clears every override (clearDeviceMetricsOverride
+ touch disabled + empty UA override) and restores the real
environment.
- Overrides are per-tab (CDP per-target) and not inherited by new tabs;
the result JSON carries a note reminding agents to re-apply.
Wire: new tool.emulate method end-to-end (bsk-protocol types + JSON
schemas + Method::ToolEmulate classified BrowserMutation, daemon
forwarding, CLI subcommand, extension tool handler + ChromiumCdp
Emulation wrappers, dispatcher registration), docs in both SKILL.md
copies, and unit tests on every layer.
* refine(emulate): merge per-tab state and align preset override semantics
- extension: remember the applied emulation state per tab and merge each new request onto it field by field before applying, so a later --width/--height no longer resets an earlier preset's dpr/mobile; --off clears the stored state along with the CDP overrides
- extension: cdp_failed errors now note that already-applied overrides were not rolled back and can be reset with --off
- cli: validate width/height pairing and metric requirements on the merged preset+flags result, so --device iphone-14 --width 500 keeps the preset height
- cli: add --no-mobile/--no-touch (conflicting with --mobile/--touch) to turn preset mobile/touch fields back off
- docs: describe the merge behaviour, --no-mobile/--no-touch and --off state clearing in both SKILL.md copies
- `bsk session start --width <px> --height <px>` opens the Agent Window
with the given outer size (chrome.windows.create)
- new `bsk window resize --session <id> --width <px> --height <px>`
resizes the Agent Window at runtime (chrome.windows.update)
- both validated on CLI, daemon and extension sides (100..=7680 CSS px)
- protocol: new tool.window_resize method (BrowserMutation) and
optional width/height on tool.session_start params; schemas regenerated
`bsk install-skill --all` installed the bundled skill into every supported harness directory (detected or not), creating junk paths like ~/.claude/skills on machines without those harnesses. --all now targets detected harnesses only and errors with guidance when none are found; --harness remains the explicit override. The --yes non-interactive branch shares the same detected-only logic.
* feat(request-help): add BSK_REQUEST_HELP=off to disable blocking help requests
* style(extension): format HelpOutcome union per biome
* docs(skill): condense BSK_REQUEST_HELP section
This PR brings the new VOM feature and greatly improves BrowserSkill. Some strategies may need user feedback to catch edge cases, so let's merge it now as a new version and continue polishing it iteratively.
