EV

event-template-mapping

Map visual detection data into structured event monitoring results.

Install

mkdir -p .claude/skills/event-template-mapping && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/16302" && unzip -o skill.zip -d .claude/skills/event-template-mapping && rm skill.zip

Installs to .claude/skills/event-template-mapping

Activation

This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.

Map LLM visual observations or method-level event results into final video monitoring Event Candidate semantics. Use when Codex has frame observations, visual timeline, template_name, template config, method_results, camera_id, score_policy, or review_threshold and needs event_type, confidence, severity, reason, evidence frames, and review flags.
348 chars✓ has a “when” triggerlonger than Claude Code's old 250-char listing cap (fine on current versions)
Intermediate

Key capabilities

  • Map LLM visual observations to event candidates
  • Translate method-level event results to event candidates
  • Generate a reason for the event from observations
  • Set confidence based on evidence quality and consistency
  • Flag events requiring review based on ambiguity or sparsity

How it works

The skill maps visual observations or method results to event candidates by applying rules based on the visual timeline and event template preconditions. It computes confidence and review status.

Inputs & outputs

You give it
frame observations, visual timeline, template_name, template config, method_results, camera_id, score_policy, review_threshold
You get back
event_type, confidence, severity, reason, evidence frames, and review flags

When to use event-template-mapping

  • Map visual observations
  • Identify event candidates
  • Set event monitoring semantics

About this skill

Event Template Mapping

Execution Priority

Prioritize mapping visual observations into event candidates. Call scripts/run.py only when structured method_results already exist or the user explicitly asks for template-rule execution.

Input Resolution

If method_results are missing but frame observations are available, map the visual observations directly into event candidates. If both visual observations and structured method results are missing, first extract and review frames through analyze-video or frame-sampling, then call the relevant method-level event skill instructions. Do not invent final event semantics from the user request alone.

LLM Visual Mapping Rules

  1. Use the visual timeline as the primary evidence source.
  2. Select the event template whose preconditions are visibly satisfied.
  3. Generate reason from concrete observations: subjects, action/state, location, timestamps, and frame ids.
  4. Set confidence from evidence quality, temporal consistency, and template fit.
  5. Set requires_review=true when the event is serious but visually ambiguous, when frames are sparse, or when only aftermath is visible.

Atomic CLI

Run this skill directly with its own script. The script does not call other skill scripts and does not depend on shared src, tools, or registry modules.

python event-template-mapping/scripts/run.py --method-results-json <method_results.json> --template-name <template_name> --camera-id <camera_id> --config <config.json> --output <event.json>

Parameters: --method-results-json, --template-json, --template-name, --camera-id, --config, --output.

Workflow

  1. Read template name, join policy, score policy, visual observations, and method results.
  2. Check whether the visible evidence satisfies required event semantics.
  3. Compute event confidence and review status.
  4. Return a standard Event Candidate schema with evidence frame ids and visual reason.

Available Implementation

This skill is implemented as an atomic standalone script in its own scripts/run.py. The script contains the executable logic for this skill and must not import shared src, tools, or registry modules.

Inputs

  • template_name
  • template
  • method_results
  • camera_id

Outputs

Returns matched=true plus Event Candidate fields, or matched=false.

Failure Modes

Returns matched=false when join policy fails or score is below score_policy.min_score.

Constraints

Do not generate evidence files. Do not map an event type when the frame evidence does not visibly support it.

When not to use it

  • When the user wants to generate evidence files
  • When the frame evidence does not visibly support an event type
  • When `method_results` and frame observations are both missing without prior frame extraction

Limitations

  • It does not generate evidence files.
  • It cannot map an event type if frame evidence does not visibly support it.
  • It returns `matched=false` when join policy fails or score is below `score_policy.min_score`.

How it compares

This skill automates the translation of raw visual or method data into structured event candidates with confidence and review flags, rather than requiring manual interpretation and categorization of video monitoring inputs.

Compared to similar skills

event-template-mapping side by side with the closest alternatives in the catalog.

SkillInstallsUpdatedSafetyDifficulty
event-template-mapping (this skill)02moReviewIntermediate
video-comparer19moReviewIntermediate
smartest-tv04moReviewBeginner
grafana-dashboards1345moNo flagsIntermediate

Try saying

Example prompts that trigger this skill in your AI assistant.

You might also like

video-comparer

daymade

This skill should be used when comparing two videos to analyze compression results or quality differences. Generates interactive HTML reports with quality metrics (PSNR, SSIM) and frame-by-frame visual comparisons. Triggers when users mention "compare videos", "video quality", "compression analysis", "before/after compression", or request quality assessment of compressed videos.

15

smartest-tv

Hybirdss

Control a smart TV or open streaming content in the browser. Play Netflix episodes/movies, YouTube videos, Spotify music — on TV or in browser (--tv browser). Also handles volume, power, notifications, history, 'continue watching', scenes, recommendations, cast URLs, queue, and multi-TV management.

00

grafana-dashboards

wshobson

Create and manage production Grafana dashboards for real-time visualization of system and application metrics. Use when building monitoring dashboards, visualizing metrics, or creating operational observability interfaces.

134441

dashboard-creator

mhattingpete

Create HTML dashboards with KPI metric cards, bar/pie/line charts, progress indicators, and data visualizations. Use when users request dashboards, metrics displays, KPI visualizations, data charts, or monitoring interfaces.

31157

model-usage

openclaw

Use CodexBar CLI local cost usage to summarize per-model usage for Codex or Claude, including the current (most recent) model or a full model breakdown. Trigger when asked for model-level usage/cost data from codexbar, or when you need a scriptable per-model summary from codexbar cost JSON.

548

analytics-tracking

davila7

When the user wants to set up, improve, or audit analytics tracking and measurement. Also use when the user mentions "set up tracking," "GA4," "Google Analytics," "conversion tracking," "event tracking," "UTM parameters," "tag manager," "GTM," "analytics implementation," or "tracking plan." For A/B test measurement, see ab-test-setup.

736

Search skills

Search the agent skills registry