agent-desktop-ffi
Skill4.8k starsChanged 2 days agoApache-2.0View on GitHub
C-ABI bindings over agent-desktop's PlatformAdapter. Consumers (Python ctypes, Swift, Node ffi-napi, Go cgo, C++, Ruby fiddle) link libagent_desktop_ffi.{dylib,so,dll} and call `ad_*` functions directly instead of spawning the CLI binary per call. The canonical observe-act workflow is: ad_init → ad_adapter_create[_with_session] → ad_snapshot → read qualified refs (@s8f3k2p9:e5) → ad_execute_by_ref → ad_free_string → ad_adapter_destroy.
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 0 are distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 68
ABI handshake: Call ad_init(AD_ABI_VERSION_MAJOR) once after loading the library.
- Very specific: line 93
Generation-safe direct APIs: Every direct targeting function takes process-generation evidence: AdExactRefEntry, AdExactWindowInfo, …
- Very specific: line 118
Display discovery: Call ad_list_displays before using AdScreenshotTarget.screen_index.
≈ “Common” is an estimate: at least 1 in 5 of the 308,232 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
agent-desktop-ffi
Direct C-ABI access to PlatformAdapter operations. Build the cdylib with the
release-ffi profile:
cargo build --profile release-ffi -p agent-desktop-ffiThe output is target/release-ffi/libagent_desktop_ffi.dylib (.so on Linux,
agent_desktop_ffi.dll on Windows) plus a committed C header at
crates/ffi/include/agent_desktop.h. The default release profile uses
panic = "abort": a Rust panic inside an extern "C" function would SIGABRT
the host. release-ffi keeps panic = "unwind", which the panic traps need.
Four reference topics, loaded as needed:
- ownership.md: who allocates and who frees every pointer the FFI hands back.
- error-handling.md: errno-style last-error contract, error codes, enum validation, panic boundary.
- threading.md: host-thread contract, mutation ordering across threads and processes, adapter-bound native handles, AXIsProcessTrusted inheritance.
- build-and-link.md: prebuilt archives, ABI handshake, struct size validation, C and Python examples.
Observe-act workflow (canonical path)
ad_init(AD_ABI_VERSION_MAJOR) // verify header ↔ dylib match
adapter = ad_adapter_create_with_session("s1") // or ad_adapter_create()
rc = ad_snapshot(adapter, "Finder", 0, 10, false, false, &json_out)
// parse json_out: locate snapshot-qualified refs in data.tree
ad_free_string(json_out)
// build action:
AdAction act = {0}; act.kind = AD_ACTION_KIND_CLICK;
rc = ad_execute_by_ref(adapter, "@s8f3k2p9:e5", &act, 0, &result_out)
ad_free_string(result_out)
ad_adapter_destroy(adapter)ad_snapshot returns the CLI envelope {version, ok, command, data}.
data.tree carries snapshot-qualified refs for interactive elements. Read
data.complete before trusting a tree: false means the observation budget ran
out and the tree is partial, not an error. ad_snapshot always targets the
focused window of app; window targeting, skeleton mode and --root drill-down
are not exposed over the ABI. A bare @eN ref returns
AD_RESULT_ERR_INVALID_ARGS.
Core constraints
- ABI handshake. Call
ad_init(AD_ABI_VERSION_MAJOR)once after loading the library. A mismatch returnsAD_RESULT_ERR_INVALID_ARGS: abort.ad_abi_version()returns the library's major for display. The major increments on a removed symbol or a layout change; newad_*symbols and error codes are additive. Pin the exactlibagent_desktop_ffiversion you link against. - Session adapters.
ad_adapter_create_with_session("session-id")scopes refmap persistence to a session namespace, like CLI--session <id>. IDs are 1 to 64 ASCII alphanumeric,-, or_characters. A null ID behaves likead_adapter_create(). An invalid ID returns null (readad_last_error_*). - Ref actions.
ad_execute_by_refis the primary action entrypoint. It loads the RefStore, resolves the ref strictly (STALE_REFwhen the snapshot has no such ref or the element changed,AMBIGUOUS_TARGETon several candidates), runs the live actionability preflight (visible, stable, enabled, supported action, policy, editable), then dispatches. A disabled or unsupported target fails before any platform call; onAD_RESULT_ERR_ACTION_FAILEDthe check report is JSON inad_last_error_details(). It auto-waits 5000 ms before the preflight;ad_execute_by_ref_timeouttakestimeout_ms(-1= 5000 ms default,0= single-shot).PressKeydefaults tofocus_fallback; every other action, includingTypeText, defaults toheadless. PassAD_POLICY_KIND_FOCUS_FALLBACK(1) to allow focus without cursor movement, orAD_POLICY_KIND_HEADED(2) to allow physical cursor and keyboard delivery.ad_execute_ref_action_exact_with_policyruns the same preflight from anAdExactRefEntry. - Generation-safe direct APIs. Every direct targeting function takes
process-generation evidence:
AdExactRefEntry,AdExactWindowInfo,ad_list_windows_exact,ad_launch_app_exact, and the other*_exactsymbols.AdRefEntry,AdWindowInfo, andAdSurfaceInfoexist only as fields embedded in the exact structs.ad_list_surfaces_exactkeeps the surfaceid. FromAD_EXACT_SURFACE_INFO_VERSION2,AdExactSurfaceInfo.unclassifiednames the surface kinds its probe could not read (for examplesheet,menu), or is null. A missingsheetormenunext to a listed kind is unknown, not absent. Only the Windows adapter fills it; macOS and Linux return null. - Native handles.
ad_resolve_element_exactandad_find_exactreturn anAdNativeHandle.ad_execute_action(headless, no preflight) andad_execute_action_with_policyact on it directly, bypassing the ref-action pipeline. Release every handle withad_free_handle(adapter, &handle)on the same adapter and thread, beforead_adapter_destroy, on every platform. It zeroeshandle.ptr, so a second call is a no-op. - Action results.
AdActionResult.stepsmirrors the CLIstepsarray for activation-chain actions. EachAdActionStephaslabelandoutcomestrings,mechanism(AD_STEP_MECHANISM_SEMANTIC_API= 1 orAD_STEP_MECHANISM_PHYSICAL_SYNTHETIC= 2, valid only whenhas_mechanism), andverified(valid only whenhas_verified).AdActionResult.dispositioncarries the same delivery/retry pair asad_last_error_delivery_semantics. Release the result withad_free_action_result(&out). - Display discovery. Call
ad_list_displaysbefore usingAdScreenshotTarget.screen_index. List order is the screenshot index order. EachAdDisplayInfocarries a stable display ID, bounds, primary flag, and scale. Release the list withad_display_list_free. - Errors. A failing call sets thread-local last-error state; read it with the
ad_last_error_*accessors before the next failing call.ad_last_error_details()returns structured JSON (the actionability report onACTION_FAILED, candidate summaries onAMBIGUOUS_TARGET) that can contain element names, values, and window titles from the user's screen: keep it out of shared logs.ad_last_error_delivery_semanticsreports whether the failed action was delivered and whether a retry is safe. Invalid enum discriminants returnAD_RESULT_ERR_INVALID_ARGS. See error-handling.md. - Threading. Entrypoints may be called from any host thread. Native handles stay on the adapter and thread that created them; pass qualified refs across threads. Desktop mutations take the same cross-process interaction lease as the CLI; reads take no lease and carry finite deadlines. See threading.md for the read/read, read/mutation, and mutation/mutation matrix.
- Tracing. File tracing (same JSONL contract as CLI
--trace; segments carryevent,ts_ms,seq, and redacted fields) is active only when the session has a manifest withtrace: on, created bysession start(or an equivalent manifest on disk) before the adapter is created.ad_adapter_create_with_sessionalone writes nothing. Each OS process appends to one segment,~/.agent-desktop/sessions/<id>/trace/<pid>-<procTs>.jsonl, and a long-lived host reuses it.ad_trace_show(JSON envelope) andad_trace_export(self-contained HTML file) read the adapter session's merged timeline;limit-1= CLI default. - Log callback.
ad_set_log_callback(cb)registers one process-wide callback for diagnostics regardless of session manifest. It receives, as JSON, thetracingevents emitted inside laterad_*calls on the same thread; the FFI never installs or replaces the host process's global tracing subscriber.cbreceives anint32_tlevel (1=ERROR ... 5=TRACE) and aconst char *msgvalid only for the duration of the call. PassNULLto unregister. Sensitive field values (password, token, text, ...) are replaced with{"redacted":true}. The callback must not unwind across the C ABI: catch C++ exceptions and Rust panics inside it, because a violation may abort the host. It may fire from several host threads at once and briefly after aNULLunregister: keep the callback and any data it captures valid for the process lifetime. - Wait.
ad_wait(adapter, args, &out)runs the CLIwaitcommand. Setmode.pause(AdOptionalU64, ms),mode.element,mode.window(waits for a window with that title),mode.text, or one ofmode.surfaces.menu,.menu_closed,.notification; add thepredicategroup (predicate,value,action,count) andscope.timeout_ms/scope.app. The CLI event wait (--event,--window-id) is not exposed. Zero-initializeAdWaitArgsand validate the struct size againstAD_WAIT_ARGS_SIZE/ad_wait_args_size()before calling. The output is a{version, ok, command, data}envelope freed withad_free_string.ad_waitblocks the calling thread up toscope.timeout_msms. A concurrentad_adapter_destroydoes not invalidate an in-flight wait. - Text input privacy. On macOS, focus-fallback or headed text insertion may
briefly use the clipboard for non-ASCII text. For sensitive text, prefer
AD_ACTION_KIND_SET_VALUEwithAD_POLICY_KIND_HEADLESSwhen the target supports settable values. ad_get_tree_exactvsad_snapshot.ad_get_tree_exactreturns a raw flat breadth-first tree: no refs, no refmap persistence, no JSON envelope (free it withad_free_tree). Use it for custom traversal or UI inspection. For observe-act agents that drive actions viaad_execute_by_ref, start withad_snapshot.
Other versions2 projects use a version of this.
Word for word · 0
None found.
Edited · 2
- lahfir/agent-desktop★ 1.8k99% same
- ferris-joy/axiomdesk★ 221% same
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository9 other files it gives its agents.
- AGENTS.mdAGENTS.md
- apps/docs/AGENTS.mdAGENTS.md
- architecture-diagramSkill
- technical-writerSkill
- agent-desktop-linuxSkill
- agent-desktop-macosSkill
- agent-desktopSkill
- agent-desktop-windowsSkill
- jev-desktopSkill
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

