feat: screen intelligence pipeline + CLI + keep_screenshots config#339
Merged
senamakel merged 4 commits intotinyhumansai:mainfrom Apr 6, 2026
Merged
Conversation
- Introduced a new `keep_screenshots` option in the Screen Intelligence settings, allowing users to save captured screenshots to the workspace instead of deleting them after processing. - Updated relevant components and tests to reflect this new feature, ensuring consistent behavior across the application. - Enhanced the user interface to include a checkbox for the `keep_screenshots` setting, improving user experience and configurability.
Adds a dedicated CLI (mirroring `openhuman skills`) to run the screen intelligence capture + vision pipeline without the full desktop app. Subcommands: run, status, capture, start, stop. Makes save_screenshot_to_disk public so the CLI capture --keep flag works. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Contributor
|
Warning Rate limit exceeded
Your organization is not enrolled in usage-based pricing. Contact your admin to enable usage-based pricing to continue reviews beyond the rate limit, or try again in 12 minutes and 27 seconds. ⌛ How to resolve this issue?After the wait time has elapsed, a review can be triggered using the We recommend that you space out your commits to avoid hitting the rate limit. 🚦 How do rate limits work?CodeRabbit enforces hourly rate limits for each developer per organization. Our paid plans have higher rate limits than the trial, open-source and free plans. In all cases, we re-allow further reviews after a brief timeout. Please see our FAQ for further information. ℹ️ Review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (1)
📝 WalkthroughWalkthroughThis pull request introduces a new Changes
Sequence Diagram(s)sequenceDiagram
participant User as User
participant UI as ScreenIntelligencePanel
participant Config as Config System
participant Engine as Screen Intelligence Engine
participant Disk as Disk Storage
User->>UI: Toggle keep_screenshots checkbox
UI->>Config: Save setting via openhumanUpdateScreenIntelligenceSettings
Config->>Config: Apply patch to config.screen_intelligence.keep_screenshots
rect rgba(200, 150, 255, 0.5)
Note over Engine,Disk: On each frame capture
Engine->>Disk: save_screenshot_to_disk(frame)
Disk-->>Engine: Return file path
Engine->>Engine: analyze_frame_with_vision(frame)
alt keep_screenshots == true
Note over Engine,Disk: Retain screenshot
Engine->>Disk: Screenshot persisted
else keep_screenshots == false
Engine->>Disk: Delete saved screenshot file
Disk-->>Engine: Deletion complete
end
end
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~22 minutes Possibly related PRs
Poem
🚥 Pre-merge checks | ✅ 2 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Actionable comments posted: 6
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (1)
src/openhuman/screen_intelligence/engine.rs (1)
906-929:⚠️ Potential issue | 🟠 MajorPotential screenshot leak when session disappears mid-loop.
Line 909 can write a screenshot before Line 924 checks session existence. If session is already gone, the
breakat Line 928 skips the deletion branch (Line 935+), so files may remain even whenkeep_screenshots=false.Suggested fix (cleanup before early break)
{ let mut state = self.inner.lock().await; if let Some(session) = state.session.as_mut() { session.vision_state = "processing".to_string(); } else { + if !keep_screenshots { + if let Some(path) = saved_path.as_ref() { + let _ = std::fs::remove_file(path); + } + } tracing::debug!("[screen_intelligence] vision worker: no session, exiting"); break; } }Also applies to: 934-944
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/openhuman/screen_intelligence/engine.rs` around lines 906 - 929, A screenshot can be saved via save_screenshot_to_disk before checking whether the session exists (self.inner.lock().await -> session), so if session is gone we break out and skip the later deletion branch causing a leak; modify the loop to, immediately after saving (i.e. where saved_path is set), check whether state.session is Some and if not then: if saved_path.is_some() and keep_screenshots is false delete the saved file (using workspace_dir and the saved_path) before breaking, and make the same change for the later mirror block (the block around session.vision_state update and subsequent deletion logic) so any saved screenshot is cleaned up when the session disappears.
🧹 Nitpick comments (1)
app/src/components/settings/panels/__tests__/ScreenIntelligencePanel.test.tsx (1)
188-194: Assertkeep_screenshotsin the save payload.Line 188 currently doesn’t verify the new flag, so this test can still pass if
keep_screenshotsis accidentally dropped during save.Suggested test assertion update
expect(openhumanUpdateScreenIntelligenceSettings).toHaveBeenCalledWith({ enabled: true, policy_mode: 'all_except_blacklist', baseline_fps: 1, + keep_screenshots: false, allowlist: ['Code'], denylist: ['1Password'], });Based on learnings: "Applies to app/src/**/*.test.{ts,tsx} : Prefer testing behavior over implementation details; use existing helpers from
app/src/test/(test-utils.tsx, shared mock backend) before adding new harness code".🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@app/src/components/settings/panels/__tests__/ScreenIntelligencePanel.test.tsx` around lines 188 - 194, Update the test assertion to include the new keep_screenshots flag in the payload so the save action is validated for that field; specifically modify the expect for openhumanUpdateScreenIntelligenceSettings to include keep_screenshots: true (or the appropriate boolean) alongside enabled, policy_mode, baseline_fps, allowlist, and denylist in the object so the test fails if keep_screenshots is omitted or changed.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@app/src/components/settings/panels/ScreenIntelligencePanel.tsx`:
- Around line 285-289: The new JSX block in the ScreenIntelligencePanel
component (the <label> block inside ScreenIntelligencePanel.tsx) is not
formatted to project Prettier rules; run Prettier over
app/src/components/settings/panels/ScreenIntelligencePanel.tsx (or the whole
app/src/**/*.{ts,tsx}) to fix formatting, then re-run lint/type checks (ESLint
and tsc --noEmit) and commit the formatted changes so `prettier --check` passes.
In `@src/core/cli.rs`:
- Around line 36-38: The new top-level command "screen-intelligence" was added
via run_screen_intelligence_command but isn’t listed in the global help output;
update the top-level help/usage text (the function or constant that prints
command listings for openhuman) to include an entry for "screen-intelligence"
with a short one-line description so it appears when users run `openhuman
--help`; locate the same help-generation place where other commands are listed
and add the "screen-intelligence" entry next to those commands so help stays
consistent with the match arm that calls
crate::core::screen_intelligence_cli::run_screen_intelligence_command.
In `@src/core/screen_intelligence_cli.rs`:
- Around line 307-415: The stop subcommand (run_stop_session) spins up a fresh
local engine via bootstrap_engine and thus only stops sessions in that process;
change it to call the global/persistent engine control API instead of creating a
new in-process engine. Replace the bootstrap_engine/engine.stop_session flow in
run_stop_session with the client/IPC path used to control the long-lived service
(e.g., use the EngineClient or connect_to_engine_daemon function used
elsewhere), so run_stop_session issues stop_session over the daemon IPC/RPC
channel (keep init_quiet_logging(opts.verbose) for logging and still call
engine_client.stop_session(Some("cli_stop".to_string())) and print
session.stop_reason as before).
- Around line 164-169: Run rustfmt (cargo fmt) to fix formatting mismatches in
this file: reformat the block that defines bind_addr, listener, and the
log::info! call (the variables bind_addr and listener and the log::info!
invocation around the JSON-RPC readiness message) so it matches project style;
after formatting, run cargo check to ensure no other style-related edits are
required and commit the formatted changes.
In `@src/openhuman/screen_intelligence/engine.rs`:
- Around line 863-870: The modified block around building app_slug and filename
is not rustfmt-compliant; run `cargo fmt` (and then `cargo check`) and commit
the reformatted changes so the `app_name` handling and the `filename` formatting
match project style, ensuring the closure using `.chars().map(...)` and the
`format!` call in the function that constructs `app_slug`/`filename` are
reformatted by rustfmt.
- Around line 859-874: The blocking std::fs calls in
run_vision_worker/save_screenshot_to_disk (notably std::fs::create_dir_all,
std::fs::write and later std::fs::remove_file) must be converted to non-blocking
async equivalents or run on a blocking thread; either replace them with
tokio::fs::create_dir_all / tokio::fs::write / tokio::fs::remove_file and await
them, or wrap the existing std::fs calls in tokio::task::spawn_blocking and
await the JoinHandle; update the code that constructs screenshots_dir,
file_path, and the write/remove operations in the functions handling screenshot
persistence (e.g., run_vision_worker and save_screenshot_to_disk) to use one of
these approaches so the Tokio runtime threads are not blocked.
---
Outside diff comments:
In `@src/openhuman/screen_intelligence/engine.rs`:
- Around line 906-929: A screenshot can be saved via save_screenshot_to_disk
before checking whether the session exists (self.inner.lock().await -> session),
so if session is gone we break out and skip the later deletion branch causing a
leak; modify the loop to, immediately after saving (i.e. where saved_path is
set), check whether state.session is Some and if not then: if
saved_path.is_some() and keep_screenshots is false delete the saved file (using
workspace_dir and the saved_path) before breaking, and make the same change for
the later mirror block (the block around session.vision_state update and
subsequent deletion logic) so any saved screenshot is cleaned up when the
session disappears.
---
Nitpick comments:
In
`@app/src/components/settings/panels/__tests__/ScreenIntelligencePanel.test.tsx`:
- Around line 188-194: Update the test assertion to include the new
keep_screenshots flag in the payload so the save action is validated for that
field; specifically modify the expect for
openhumanUpdateScreenIntelligenceSettings to include keep_screenshots: true (or
the appropriate boolean) alongside enabled, policy_mode, baseline_fps,
allowlist, and denylist in the object so the test fails if keep_screenshots is
omitted or changed.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: bf5a01cf-9078-4caa-bb13-380e19e2c2ff
📒 Files selected for processing (14)
app/src/components/intelligence/__tests__/ScreenIntelligenceDebugPanel.test.tsxapp/src/components/settings/panels/ScreenIntelligencePanel.tsxapp/src/components/settings/panels/__tests__/AccessibilityPanel.test.tsxapp/src/components/settings/panels/__tests__/ScreenIntelligencePanel.test.tsxapp/src/services/__tests__/coreRpcClient.test.tsapp/src/store/__tests__/accessibilitySlice.test.tsapp/src/utils/tauriCommands.tssrc/core/cli.rssrc/core/mod.rssrc/core/screen_intelligence_cli.rssrc/openhuman/config/ops.rssrc/openhuman/config/schema/accessibility.rssrc/openhuman/config/schemas.rssrc/openhuman/screen_intelligence/engine.rs
Comment on lines
+285
to
+289
| <label className="flex items-center justify-between rounded-xl border border-stone-200 bg-stone-50 px-3 py-2"> | ||
| <div> | ||
| <span className="text-sm text-stone-700">Keep Screenshots</span> | ||
| <p className="text-xs text-stone-400">Save captured screenshots to the workspace instead of deleting after processing</p> | ||
| </div> |
Contributor
There was a problem hiding this comment.
Run Prettier on the new JSX block before merge.
The pipeline reports prettier --check failure, and this newly added block is in the failing file.
As per coding guidelines: app/src/**/*.{ts,tsx}: Run Prettier, ESLint, and tsc --noEmit before merging changes to app/ code.
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@app/src/components/settings/panels/ScreenIntelligencePanel.tsx` around lines
285 - 289, The new JSX block in the ScreenIntelligencePanel component (the
<label> block inside ScreenIntelligencePanel.tsx) is not formatted to project
Prettier rules; run Prettier over
app/src/components/settings/panels/ScreenIntelligencePanel.tsx (or the whole
app/src/**/*.{ts,tsx}) to fix formatting, then re-run lint/type checks (ESLint
and tsc --noEmit) and commit the formatted changes so `prettier --check` passes.
Comment on lines
+36
to
+38
| "screen-intelligence" => { | ||
| crate::core::screen_intelligence_cli::run_screen_intelligence_command(&args[1..]) | ||
| } |
Contributor
There was a problem hiding this comment.
Expose the new command in general help output.
Line 36 adds a new top-level command, but openhuman --help doesn’t list it yet, which hurts discoverability.
Suggested help text update
fn print_general_help(grouped: &BTreeMap<String, Vec<ControllerSchema>>) {
println!("OpenHuman core CLI\n");
println!("Usage:");
println!(" openhuman run [--host <addr>] [--port <u16>] [--jsonrpc-only] [--verbose]");
println!(" openhuman repl [--verbose] [--eval '<cmd>'] [--batch]");
println!(" openhuman call --method <name> [--params '<json>']");
println!(" openhuman skills <subcommand> [options] (skill development runtime)");
+ println!(" openhuman screen-intelligence <subcommand> [options] (capture + vision debug)");
println!(" openhuman <namespace> <function> [--param value ...]\n");🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@src/core/cli.rs` around lines 36 - 38, The new top-level command
"screen-intelligence" was added via run_screen_intelligence_command but isn’t
listed in the global help output; update the top-level help/usage text (the
function or constant that prints command listings for openhuman) to include an
entry for "screen-intelligence" with a short one-line description so it appears
when users run `openhuman --help`; locate the same help-generation place where
other commands are listed and add the "screen-intelligence" entry next to those
commands so help stays consistent with the match arm that calls
crate::core::screen_intelligence_cli::run_screen_intelligence_command.
Comment on lines
+164
to
+169
| let bind_addr = format!("127.0.0.1:{}", opts.port); | ||
| let listener = tokio::net::TcpListener::bind(&bind_addr).await?; | ||
|
|
||
| log::info!( | ||
| "[screen-intelligence-cli] ready — http://{bind_addr}/rpc (JSON-RPC 2.0)" | ||
| ); |
Contributor
There was a problem hiding this comment.
Apply cargo fmt for this file before merge.
CI is reporting formatting mismatches in these changed blocks.
As per coding guidelines: src/**/*.rs: Run cargo fmt and cargo check before merging changes to Rust code.
Also applies to: 355-358
🧰 Tools
🪛 GitHub Actions: Type Check
[error] 164-164: cargo fmt --check failed due to formatting mismatch. Expected log::info! macro to be on a single line: log::info!("[screen-intelligence-cli] ready — http://{bind_addr}/rpc (JSON-RPC 2.0)");
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@src/core/screen_intelligence_cli.rs` around lines 164 - 169, Run rustfmt
(cargo fmt) to fix formatting mismatches in this file: reformat the block that
defines bind_addr, listener, and the log::info! call (the variables bind_addr
and listener and the log::info! invocation around the JSON-RPC readiness
message) so it matches project style; after formatting, run cargo check to
ensure no other style-related edits are required and commit the formatted
changes.
Comment on lines
+307
to
+415
| /// `openhuman screen-intelligence start` — start a capture + vision session. | ||
| fn run_start_session(args: &[String]) -> Result<()> { | ||
| if args.iter().any(|a| is_help(a)) { | ||
| println!("Usage: openhuman screen-intelligence start [--ttl <secs>] [-v]"); | ||
| println!(); | ||
| println!("Start a screen intelligence capture session with vision analysis."); | ||
| println!("The session runs until TTL expires or Ctrl+C is pressed."); | ||
| println!(); | ||
| println!(" --ttl <secs> Session duration (default: 300, max: 3600)"); | ||
| println!(" -v, --verbose Enable debug logging"); | ||
| return Ok(()); | ||
| } | ||
|
|
||
| let (opts, _) = parse_opts(args)?; | ||
| crate::core::logging::init_for_cli_run(opts.verbose); | ||
|
|
||
| let rt = tokio::runtime::Builder::new_multi_thread() | ||
| .enable_all() | ||
| .build()?; | ||
|
|
||
| rt.block_on(async { | ||
| let engine = bootstrap_engine(opts.verbose).await?; | ||
|
|
||
| let params = crate::openhuman::screen_intelligence::StartSessionParams { | ||
| consent: true, | ||
| ttl_secs: Some(opts.ttl_secs), | ||
| screen_monitoring: Some(true), | ||
| device_control: Some(false), | ||
| predictive_input: Some(false), | ||
| }; | ||
|
|
||
| match engine.start_session(params).await { | ||
| Ok(session) => { | ||
| eprintln!(" Session started!"); | ||
| eprintln!(" TTL: {}s", session.ttl_secs); | ||
| eprintln!(" Vision: {}", session.vision_enabled); | ||
| eprintln!(" Panic hotkey: {}", session.panic_hotkey); | ||
| eprintln!(); | ||
| eprintln!(" Capturing screenshots and running vision analysis..."); | ||
| eprintln!(" Press Ctrl+C to stop."); | ||
| eprintln!(); | ||
|
|
||
| // Print periodic status updates until the session ends. | ||
| let mut tick = tokio::time::interval(std::time::Duration::from_secs(5)); | ||
| loop { | ||
| tick.tick().await; | ||
| let status = engine.status().await; | ||
| if !status.session.active { | ||
| eprintln!( | ||
| "\n Session ended: {}", | ||
| status.session.stop_reason.unwrap_or_else(|| "unknown".into()) | ||
| ); | ||
| break; | ||
| } | ||
| eprintln!( | ||
| " [{}] captures={} vision={} queue={} last_app={:?}", | ||
| chrono::Utc::now().format("%H:%M:%S"), | ||
| status.session.capture_count, | ||
| status.session.vision_state, | ||
| status.session.vision_queue_depth, | ||
| status.session.last_context.as_deref().unwrap_or("-"), | ||
| ); | ||
| if let Some(summary) = &status.session.last_vision_summary { | ||
| let truncated = if summary.len() > 100 { | ||
| format!("{}…", &summary[..100]) | ||
| } else { | ||
| summary.clone() | ||
| }; | ||
| eprintln!(" notes: {truncated}"); | ||
| } | ||
| } | ||
| } | ||
| Err(e) => { | ||
| eprintln!(" Failed to start session: {e}"); | ||
| std::process::exit(1); | ||
| } | ||
| } | ||
| Ok(()) | ||
| }) | ||
| } | ||
|
|
||
| /// `openhuman screen-intelligence stop` — stop an active session. | ||
| fn run_stop_session(args: &[String]) -> Result<()> { | ||
| if args.iter().any(|a| is_help(a)) { | ||
| println!("Usage: openhuman screen-intelligence stop [-v]"); | ||
| println!(); | ||
| println!("Stop the active screen intelligence session."); | ||
| return Ok(()); | ||
| } | ||
|
|
||
| let (opts, _) = parse_opts(args)?; | ||
| init_quiet_logging(opts.verbose); | ||
|
|
||
| let rt = tokio::runtime::Builder::new_multi_thread() | ||
| .enable_all() | ||
| .build()?; | ||
|
|
||
| rt.block_on(async { | ||
| let engine = bootstrap_engine(opts.verbose).await?; | ||
| let session = engine.stop_session(Some("cli_stop".to_string())).await; | ||
| eprintln!( | ||
| " Session stopped: {}", | ||
| session | ||
| .stop_reason | ||
| .unwrap_or_else(|| "no active session".into()) | ||
| ); | ||
| Ok(()) | ||
| }) | ||
| } |
Contributor
There was a problem hiding this comment.
stop subcommand is process-local and won’t stop sessions started elsewhere.
start keeps the session in the current process, while stop (Line 405+) boots a fresh engine in a new process and stops that local state only. In practice, it won’t stop a session started by a different CLI process.
🧰 Tools
🪛 GitHub Actions: Type Check
[error] 354-354: cargo fmt --check failed due to formatting mismatch. stop_reason unwrap_or_else expression formatting differs (line wrapping/indentation).
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@src/core/screen_intelligence_cli.rs` around lines 307 - 415, The stop
subcommand (run_stop_session) spins up a fresh local engine via bootstrap_engine
and thus only stops sessions in that process; change it to call the
global/persistent engine control API instead of creating a new in-process
engine. Replace the bootstrap_engine/engine.stop_session flow in
run_stop_session with the client/IPC path used to control the long-lived service
(e.g., use the EngineClient or connect_to_engine_daemon function used
elsewhere), so run_stop_session issues stop_session over the daemon IPC/RPC
channel (keep init_quiet_logging(opts.verbose) for logging and still call
engine_client.stop_session(Some("cli_stop".to_string())) and print
session.stop_reason as before).
Comment on lines
+859
to
+874
| let screenshots_dir = workspace_dir.join("screenshots"); | ||
| std::fs::create_dir_all(&screenshots_dir) | ||
| .map_err(|e| format!("failed to create screenshots dir: {e}"))?; | ||
|
|
||
| let app_slug = frame | ||
| .app_name | ||
| .as_deref() | ||
| .unwrap_or("unknown") | ||
| .chars() | ||
| .map(|c| if c.is_alphanumeric() || c == '-' { c } else { '_' }) | ||
| .collect::<String>(); | ||
| let filename = format!("{}_{}.png", frame.captured_at_ms, app_slug); | ||
| let file_path = screenshots_dir.join(&filename); | ||
|
|
||
| std::fs::write(&file_path, &raw_bytes) | ||
| .map_err(|e| format!("failed to write screenshot {filename}: {e}"))?; |
Contributor
There was a problem hiding this comment.
🛠️ Refactor suggestion | 🟠 Major
🧩 Analysis chain
🏁 Script executed:
#!/bin/bash
set -euo pipefail
echo "Locate blocking filesystem calls in engine:"
rg -n 'std::fs::(create_dir_all|write|remove_file)' src/openhuman/screen_intelligence/engine.rs
echo
echo "Show call context around vision worker and screenshot save:"
rg -n -A8 -B8 'save_screenshot_to_disk|run_vision_worker' src/openhuman/screen_intelligence/engine.rsRepository: tinyhumansai/openhuman
Length of output: 3926
Use tokio::fs or spawn_blocking for blocking I/O in the vision worker.
Blocking std::fs calls (create_dir_all, write, remove_file) in run_vision_worker and save_screenshot_to_disk will block Tokio worker threads under load. Convert these to use tokio::fs or wrap in tokio::task::spawn_blocking to prevent stalling other async tasks.
Also applies to: lines 934-943
🧰 Tools
🪛 GitHub Actions: Type Check
[error] 865-865: cargo fmt --check failed due to formatting mismatch. Adjusted map closure formatting to multi-line if/else.
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@src/openhuman/screen_intelligence/engine.rs` around lines 859 - 874, The
blocking std::fs calls in run_vision_worker/save_screenshot_to_disk (notably
std::fs::create_dir_all, std::fs::write and later std::fs::remove_file) must be
converted to non-blocking async equivalents or run on a blocking thread; either
replace them with tokio::fs::create_dir_all / tokio::fs::write /
tokio::fs::remove_file and await them, or wrap the existing std::fs calls in
tokio::task::spawn_blocking and await the JoinHandle; update the code that
constructs screenshots_dir, file_path, and the write/remove operations in the
functions handling screenshot persistence (e.g., run_vision_worker and
save_screenshot_to_disk) to use one of these approaches so the Tokio runtime
threads are not blocked.
Comment on lines
+863
to
+870
| let app_slug = frame | ||
| .app_name | ||
| .as_deref() | ||
| .unwrap_or("unknown") | ||
| .chars() | ||
| .map(|c| if c.is_alphanumeric() || c == '-' { c } else { '_' }) | ||
| .collect::<String>(); | ||
| let filename = format!("{}_{}.png", frame.captured_at_ms, app_slug); |
Contributor
There was a problem hiding this comment.
cargo fmt needs to be applied to this changed block.
CI reports a formatting mismatch in this region.
As per coding guidelines: src/**/*.rs: Run cargo fmt and cargo check before merging changes to Rust code.
🧰 Tools
🪛 GitHub Actions: Type Check
[error] 865-865: cargo fmt --check failed due to formatting mismatch. Adjusted map closure formatting to multi-line if/else.
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@src/openhuman/screen_intelligence/engine.rs` around lines 863 - 870, The
modified block around building app_slug and filename is not rustfmt-compliant;
run `cargo fmt` (and then `cargo check`) and commit the reformatted changes so
the `app_name` handling and the `filename` formatting match project style,
ensuring the closure using `.chars().map(...)` and the `format!` call in the
function that constructs `app_slug`/`filename` are reformatted by rustfmt.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This was referenced Apr 6, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
{workspace_dir}/screenshots/, processed by local Ollama vision model, then cleaned up (or kept based on config)keep_screenshotsconfig flag (default:false): When enabled, screenshots persist on disk after vision processing instead of being deletedopenhuman screen-intelligenceCLI: Standalone CLI for testing the capture + vision pipeline without the full desktop app (mirrorsopenhuman skillspattern)CLI subcommands
screen-intelligence runscreen-intelligence statusscreen-intelligence capture [--keep]screen-intelligence start [--ttl <secs>]screen-intelligence stopConfig flag
Files changed
accessibility.rs,ops.rs,schemas.rs— addedkeep_screenshotsfieldengine.rs—save_screenshot_to_disk()+ vision worker integrationscreen_intelligence_cli.rs, wired intocli.rs+mod.rstauriCommands.ts,ScreenIntelligencePanel.tsx— UI toggle + type updateskeep_screenshots: falseTest plan
cargo checkpassestsc --noEmitpassesopenhuman screen-intelligence --helpprints usageopenhuman screen-intelligence statusprints JSON with permissions, session, configopenhuman screen-intelligence capture --keepsaves screenshot to~/.openhuman/workspace/screenshots/openhuman screen-intelligence start --ttl 60 -vruns capture + vision loopkeep_screenshots = false🤖 Generated with Claude Code
Summary by CodeRabbit
New Features
Tests