browser-interaction
Skill10 starsChanged 2 months agoApache-2.0View on GitHub
Drive live page interactions with production-grade clicks (normal/double/right/human-like), typing modes, hover/focus, keyboard shortcuts, drag-and-drop, form controls and condition-based waits.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- MatanelGordon/skillstheir own version, first seen Aug 21, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- MatanelGordon/skills★ 0same name
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 repository16 other files it gives its agents.
- browser-forensicsSkill
- browser-state-recoverySkill
- dom-inspectionSkill
- dom-mutationSkill
- dom-targetingSkill
- extension-troubleshootingSkill
- human-like-interactionSkill
- mcp-troubleshootingSkill
- page-annotationSkill
- project-captureSkill
- project-reconstructionSkill
- regression-testingSkill
- selector-recoverySkill
- tab-managementSkill
- viewport-testingSkill
- visual-dom-comparisonSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 31
Read every InteractionResult: beforeState / afterState, events fired, and for clicks which mode ran.
- Distinctive: line 54
Every result carries before/after state — compare them.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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
Skill: Browser Interaction
PURPOSE
Interact with elements the way real automation should: explicit modes that never silently fall back, before/after state on every result, and waits tied to observable conditions instead of arbitrary sleeps.
WHEN TO USE
- Executing any user-visible action: click, type, hover, focus, key press, drag, check, select.
- Filling forms where framework event handling matters (React/Vue need full key sequences).
- Waiting for a page to settle before the next action in a multi-step flow.
PREREQUISITES
- Live page connected (extension) or Node simulation context (
simulated: trueresults — event dispatch is real in JSDOM, layout/occlusion checks are limited). - A resolved target: selector or TARGET object. Generate one first (
generate_element_target) for anything you will reuse. - Optional: an interaction profile set via the human-like-interaction skill.
WORKFLOW
click_element {target|selector, mode, waitForStabilization}— modes:normal(synthetic pointer sequence),double(dblclick),right(contextmenu),human-like(trajectory + profile-driven delays),programmatic. The mode actually used is always reported.type_text {target|selector, text, mode}—append(default),replace(clear then type),clear. Framework-sensitive: keydown/keypress/input/change dispatched per character; the resulting value shows inafterState.hover_element,focus_element,blur_element— pointerenter/mouseenter/mouseover/mousemove sequence, native focus()/blur() plus events.press_keyboard_shortcut {keys, target?}—"Control+Shift+P"orkeyList: ["Control","Shift","P"]; keydown+keyup per key with modifier flags; defaults to the active element.drag_and_drop {source, target?, offsets?}— HTML5 drag events (dragstart/dragenter/dragover/drop/dragend) plus pointer events; useoffsets {x,y}when there is no drop target; result reports whether HTML5 DnD was used.set_input_checked {target|selector, checked}— checkbox/radio; peer radios sharing the samenameare deselected; dispatches input + change; returns checkedBefore/checkedAfter.select_option {target|selector, value}— sets the value and dispatches change.wait_for_condition {kind, ..., timeoutMs, pollIntervalMs}— kinds:dom_stable,selector_present,selector_visible,selector_absent,text_present,url_contains,element_count(needscount),readiness_state(defaultcomplete). timeoutMs default 5000, max 30000; poll 100ms.wait_for_dom_stable {timeoutMs}is the shorthand.- Read every
InteractionResult:beforeState/afterState, events fired, and for clicks which mode ran.
TOOLS
click_element, type_text, hover_element, focus_element, blur_element, press_keyboard_shortcut, scroll_to_element, scroll_page, drag_and_drop, set_input_checked, select_option, wait_for_condition, wait_for_dom_stable, interact_with_element
EXAMPLES
tools/call click_element { "selector": "button[data-testid='submit']", "mode": "double", "waitForStabilization": true }
tools/call type_text { "selector": "#email", "text": "a@b.co", "mode": "replace" }
tools/call press_keyboard_shortcut { "keys": "Control+Shift+P" }
tools/call drag_and_drop { "source": { "selector": ".card:nth-child(1)" }, "target": { "selector": ".drop-zone" } }
tools/call set_input_checked { "selector": "input[name='agree']", "checked": true }
tools/call wait_for_condition { "kind": "selector_visible", "selector": ".toast-success", "timeoutMs": 8000 }FAILURE MODES
- TARGET_NOT_FOUND / TARGET_STALE — element gone or removed between resolution and dispatch; run
wait_for_condition {kind:"dom_stable"}then retry, else use selector-recovery. - Click has no effect — element covered/occluded or a dead target: verify with
analyze_dom {analyzer: "detect_dead_click_targets"}andget_element_visual_state. - Typing produces nothing in a controlled component — the app listens on events the page synthesizes differently; try
mode: "replace"and confirm viaafterState. - Wait timeouts (satisfied: false, waitedMs = timeoutMs) — condition never became true; do not blindly raise the timeout past 30000 (the cap) — diagnose instead.
- Wrong element clicked after multi-match — an ambiguous selector got visible-first disambiguation; tighten the selector or pass a full TARGET.
VALIDATION
- Every result carries before/after state — compare them; a click with identical before/after usually means the handler did nothing.
set_input_checkedreports checkedBefore/checkedAfter and fired events; radio peers must show the switch.wait_for_conditionreturnssatisfied,waitedMs,detail—detailsays what was actually observed.
RECOVERY
- Stale target mid-flow:
recover_selectorwith the stored snapshot, re-rungenerate_element_target, retry the action. - Flaky overlays/toasts:
wait_for_condition {kind:"selector_absent", selector:".overlay"}before clicking. - For legacy flows,
interact_with_element {action, selector, text, key, optionValue, scrollDelta, waitForStabilization}still covers click/type/hover/focus/scroll/press_key/select in one dispatch.
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.

