Components › Agent Activity › Tool Execution Log
AG02 · Activity

Tool Execution Log

Purpose Shows discrete tool calls or system operations triggered by the AI — search, read, write, API calls. Accountability at the execution level.
Use when An agent has access to tools and is invoking them during a task. The tool execution log makes each tool call visible: what tool, what input, what result, how long it took.
Do not use As the primary UI for multi-step execution — that is the Agent Activity Timeline (AG01). The tool log is supplementary detail — a drill-down, not the top-level view.
Failure mode prevented An AI agent silently calls external APIs, reads files, and modifies data while the user sees only a loading indicator. Accountability requires that tool invocations are logged and inspectable.
Example Dark Academia
Dark Academia
The Atelier
Neon Tide
The Archive
Candlelight
The Conservatory
Tool Log 5 calls · 4.1s total
  • read search_records "pipeline errors last 7 days" 127 records 1.2s
  • read read_file incident-template.md 4.1 KB 0.3s
  • api salesforce_query SELECT error_type, count FROM logs 42 rows 2.1s
  • write write_file incident-report-q1.md Running… –
  • api notify_slack #incidents 403 Forbidden 0.1s
Overview
tool-log-compact — one-line entries with expand-on-click for full detailtool-log-verbose — each entry shows full input and output inlinetool-log-filtered — shows only specific tool types (e.g., write operations only)
Design rule

Write operations — file writes, API calls that modify state, database updates — must be visually distinguished from read operations. Reads are informational. Writes are consequential. The distinction must be visible at a glance.

Theming API

Override these CSS custom properties to apply your own visual treatment without touching the component's behavioral contracts.

Shared semantic tokens

Token Controls
--rad-surface-card Component card background
--rad-border-subtle Card border
--rad-color-text-secondary Tool input text
--rad-color-text-meta Result and time metadata
--rad-color-approve Focus ring
--rad-color-reject Error color fallback
--rad-radius-card Card border radius

Component-scoped tokens

Token Controls
--rad-tool-log-entry-border Row separator line
--rad-tool-log-active-bg Active entry row background
--rad-tool-log-error-bg Error entry row background
--rad-tool-log-error-color Error result text color
--rad-tool-log-name-color Tool name text color
--rad-tool-log-pending-color Running result text color

"*Every tool call is a decision the agent made on your behalf. Log them all — the ones that worked and the ones that didn't.*"

← Previous Agent Activity Timeline Next → Collapsible Agent Steps

Intellectual Property

RAD is the original work of Jackie Curry. All rights reserved. No portion may be reproduced, adapted, or incorporated into any product or system without express written permission.

Permitted: citation in academic or editorial contexts with full attribution.

Licensing

© 2025 Jackie Curry. All rights reserved. Publication date: 2025.

For licensing inquiries, connect on LinkedIn →