effort

Estimates implementation effort for issues based on code exploration output.

Updated Mar 22, 2026
One-click install
npx skills add https://github.com/bkrsna/gemini-cli --skill effort-bkrsna
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: effort
Source: https://github.com/bkrsna/gemini-cli/tree/main/tools/caretaker-agent/cloudrun/triage-worker/.gemini/skills/effort
Command: npx skills add https://github.com/bkrsna/gemini-cli --skill effort-bkrsna

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Triage teams need consistent, defensible effort estimates for incoming issues, but manual sizing is subjective and slow. This Skill analyzes an issue's title, body, and code exploration output to produce a structured SMALL, MEDIUM, or LARGE effort estimate with reasoning. ## Core Features & Use Cases - Structured Effort Classification: Outputs a JSON object with an effort_estimate (SMALL, MEDIUM, or LARGE) and detailed effort_reasoning. - Domain-Specific Sizing Rules: Applies calibrated criteria for the Gemini CLI codebase, covering Ink UI components, node-pty and signal handling, MCP transport, A2A protocol, and cross-package refactors. - Reproducibility Guardrails: Automatically excludes intermittent, platform-specific, or hard-to-reproduce bugs from the SMALL category due to testing overhead. - Use Case: During automated issue triage, feed an issue report plus discovered source files into this Skill to receive a consistent effort label that informs prioritization and sprint planning. ## Quick Start Estimate the implementation effort for this issue using the attached code exploration output and return the JSON result.

Frequently Asked Questions about effort

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I estimate engineering effort for a GitHub issue automatically?

Provide the issue title, body, and code exploration output (discovered source files, UI components, and test files) to the Skill. It returns a JSON object with an effort_estimate of SMALL, MEDIUM, or LARGE plus detailed reasoning.

What effort levels does issue triage estimation support?

Three levels are supported: SMALL for fixes of one day or less, MEDIUM for two to three days of work, and LARGE for three or more days. Each level has defined criteria covering UI tweaks, async flows, platform-specific issues, and architectural changes.

What counts as a SMALL versus MEDIUM effort estimate?

SMALL covers trivial config changes, minor Ink UI adjustments, typos, and localized single-file bug fixes. MEDIUM covers React state management issues, asynchronous flow problems, tool output parsers, and refactors spanning the cli and core packages.

Can intermittent or hard-to-reproduce bugs be rated as small effort?

No. Bugs described as intermittent, flickering, platform-specific, or requiring cross-environment setups such as IDE companions must not be rated SMALL, because reproduction and testing overhead increases the true cost.

What input does the effort estimation skill require?

It requires the issue content (title and body) plus code exploration output listing discovered source files, coupled UI components, and test files. Without the exploration context, estimates cannot account for cross-component impact.