oracle-agent

System prompt and behavioral rules for the Repository OS Oracle — the Claude-powered Q&A agent that answers questions about an indexed codebase using vector retrieval from ChromaDB.

Quality

54%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./tile-source/skills/oracle-agent/SKILL.md

Quality

Content

92%Scale 1-3

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

This is a high-quality skill that provides a complete, actionable system prompt template for the Oracle agent. The content is concise, well-structured, and includes concrete examples of tool schemas and output formats. The only weakness is a minor gap in progressive disclosure where the referenced 'spec-generator skill' lacks a clear link.

Suggestions

Add an explicit link to the spec-generator skill file (e.g., 'See [spec-generator.md](spec-generator.md) for the prompt template') instead of just mentioning it by name.

Dimension	Reasoning	Score
Conciseness	The content is lean and efficient, providing only what Claude needs to know for this specific agent configuration. No unnecessary explanations of concepts Claude already understands; every section serves a clear purpose.	3 / 3
Actionability	Provides a complete, copy-paste ready system prompt template with concrete rules, exact JSON tool schema, and specific output format examples. The behavioral rules are explicit and numbered for easy reference.	3 / 3
Workflow Clarity	Clear sequencing with explicit validation checkpoints: always retrieve before answering, flag insufficient context, acknowledge staleness. Rule 8 provides a feedback loop (up to 3 retrieval calls with rephrasing). The workflow for handling edge cases is explicit.	3 / 3
Progressive Disclosure	Content is well-organized with clear sections (System Prompt, Tool Schema, Tool Result Format, Behavioral Notes), but references to external skills like 'spec-generator skill' are mentioned without clear navigation links. The content is appropriately sized but could benefit from clearer cross-references.	2 / 3
	Total	11 / 12 Passed

Description

17%Scale 1-3

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

This description reads more like internal documentation about what the skill IS rather than guidance for when Claude should USE it. It lacks natural trigger terms users would say, provides no explicit 'Use when...' guidance, and focuses on implementation details (ChromaDB, vector retrieval) rather than user-facing capabilities.

Suggestions

Add a 'Use when...' clause with natural triggers like 'when users ask questions about the codebase', 'search for code', 'find where X is defined', or 'understand how the repository works'.

Replace technical jargon with user-facing action verbs: instead of 'vector retrieval from ChromaDB', say 'Search and answer questions about code in the repository'.

List specific capabilities users would request: 'find function definitions', 'explain code relationships', 'locate usage of specific APIs', 'answer architecture questions'.

Dimension	Reasoning	Score
Specificity	Names the domain (Q&A agent, indexed codebase) and mentions vector retrieval from ChromaDB, but doesn't list specific concrete actions like 'search code', 'explain functions', or 'find dependencies'.	2 / 3
Completeness	Describes what it is (system prompt for a Q&A agent) but lacks any 'Use when...' clause or explicit trigger guidance. The 'when' is entirely missing.	1 / 3
Trigger Term Quality	Uses technical jargon ('vector retrieval', 'ChromaDB', 'Repository OS Oracle') that users wouldn't naturally say. Missing natural terms like 'search code', 'find in codebase', 'code questions', or 'understand repository'.	1 / 3
Distinctiveness Conflict Risk	The ChromaDB and 'Repository OS Oracle' references provide some distinctiveness, but 'answers questions about a codebase' could overlap with general code explanation or documentation skills.	2 / 3
	Total	6 / 12 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 11 / 11 Passed

Validation for skill structure

No warnings or errors.

Repository: tombrewsviews/repository-os-mvp
Path: tile-source/skills/oracle-agent/SKILL.md
Commit: a30d1b4

Reviewed: ~22 hours ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.