docs: Document Auto-configure repository AI feature (OD-420)

[LipeGheno]

From 3bac6c7929c4f00b04e2976a3ada67b185447996 Mon Sep 17 00:00:00 2001
From: Claude noreply@anthropic.com
Date: Thu, 27 Aug 2026 13:00:11 +0000
Subject: [PATCH] docs: Document Auto-configure repository AI feature OD-420

Adds a new "Auto-configure repository" section to the Codacy AI page,
following the same feature-block pattern used for AI-enhanced comments,
Smart False Positive Triage, and AI Reviewer.

Covers what the feature does, what happens during a run, how to turn it
on, and operational limits (admin-only, concurrency limits, bounded run
time/usage). Several specifics (exact button/menu labels, exact plan
tier) are marked with TODO: verify comments pending confirmation against
the shipped UI and the still-open release checklist (OD-577).

This covers the customer-facing half of OD-420 only. Internal reference
material for the feature is out of scope for docs.codacy.com and needs
a separate home.

docs/codacy-ai/codacy-ai.md | 37 +++++++++++++++++++++++++++++++++++++
1 file changed, 37 insertions(+)

diff --git a/docs/codacy-ai/codacy-ai.md b/docs/codacy-ai/codacy-ai.md
index d008ab4..a8f6f01 100644
--- a/docs/codacy-ai/codacy-ai.md
+++ b/docs/codacy-ai/codacy-ai.md
@@ -126,3 +126,40 @@ Rules for the output:

• To enrich the review, the git diff of the Pull Request as well as some related files' contents can be sent as context. No data is stored on our side, or used to train any models.
• Prompts are neither stored nor visible to anyone.
• As an extra precaution, before any code snippet is sent to the AI model, Codacy automatically redacts secrets (API keys, tokens, credentials, and other high-entropy strings) from the code context.

+### Auto-configure repository {: id="auto-configure-repository"}
+
+{%

• include-markdown "../assets/includes/paid.md"
• start=""
• end=""
  +%}

+
+
+This feature uses an AI coding agent, and is strictly opt-in: it only runs when an organization admin starts it for a specific repository.
+
+Auto-configure repository sets up Codacy analysis for a repository without the manual steps in Configuring your repository. An AI agent reads the repository, then applies settings such as which tools and code patterns to enable, which paths to exclude, and how to configure each language — the same settings you could otherwise set by hand or through a Codacy configuration file.
+
+What happens during a run
+
+1. Codacy clones the repository into an isolated environment, using credentials scoped and time-limited to that one run.
+2. An AI agent reviews the repository's structure, languages, and frameworks.
+3. The agent applies a Codacy configuration based on what it finds.
+4. Codacy re-analyzes the repository using the new configuration.
+
+How to turn it on
+
+1. Go to the repository's dashboard in Codacy.
+2. Select Auto-configure repository. Only organization admins see this option.
+3. Start a run, then follow its status (queued, running, successful, or failed) from the dashboard.
+4. Once a run finishes, select Review to see a summary of what changed, including the languages and files the agent analyzed.
+
+Notes
+
+- Only organization admins can start a run.
+- Codacy limits how many runs happen at the same time, both for the organization and across Codacy, so runs stay fast for everyone. If you start a run while one is already queued or running for the same repository, Codacy queues it instead of starting it immediately.
+- Each run has a fixed amount of time and AI usage it can use, so a run always finishes or fails instead of continuing indefinitely.
+- You can start a new run at any time; it replaces the configuration from the previous run.
+- The agent only has access to the repository's code and to the credentials needed for that one run. Codacy discards both when the run ends.
+- Codacy does not use your code, repository contents, or comments to train external AI models. No customer code is incorporated into model training.

2.43.0

[github-actions]

Overall readability score: 53.83 (🟢 +0.02)

File	Readability
codacy-ai.md	59.49 (🟢 +3.99)

View detailed metrics

🟢 - Shows an increase in readability
🔴 - Shows a decrease in readability

File	Readability	FRE	GF	ARI	CLI	DCRS
codacy-ai.md	59.49	37.3	10.2	11.8	11.83	6.86
	🟢 +3.99	🟢 +0.41	🟢 +0.74	🟢 +0.5	🟢 +0.41	🟢 +0.42

Averages:

	Readability	FRE	GF	ARI	CLI	DCRS
Average	53.83	42.85	10.96	12.39	12.3	8.03
	🟢 +0.02	🟢 +0	🟢 +0	🟢 +0	🟢 +0	🟢 +0

View metric targets

Metric	Range	Ideal score
Flesch Reading Ease	100 (very easy read) to 0 (extremely difficult read)	60
Gunning Fog	6 (very easy read) to 17 (extremely difficult read)	8 or less
Auto. Read. Index	6 (very easy read) to 14 (extremely difficult read)	8 or less
Coleman Liau Index	6 (very easy read) to 17 (extremely difficult read)	8 or less
Dale-Chall Readability	4.9 (very easy read) to 9.9 (extremely difficult read)	6.9 or less

[codacy-production]

Up to standards ✅

🟢 Issues 0 issues

Results:
0 new issues

View in Codacy

AI Reviewer: first review requested successfully. AI can make mistakes. Always validate suggestions.

_{TIP This summary will be updated as you push new changes.}

[codacy-production]

Pull Request Overview

This PR introduces documentation for the 'Auto-configure repository' AI feature. While the structure and technical flow descriptions generally align with requirements and meet quality standards (passing Codacy analysis), the content is currently incomplete for production release. Specifically, there are unresolved gaps concerning final UI button labels and the assigned pricing tier, both of which are currently held by placeholders and internal TODO comments. These gaps correlate with 'missing' status in key acceptance criteria. Additionally, the document contains redundant privacy information and ambiguous section headers that should be refined for better clarity and maintainability. It is recommended to resolve the dependencies on ticket OD-577 before merging to prevent publishing inaccurate documentation.

About this PR

• Completion of this documentation is blocked by the finalization of UI labels and pricing tiers in OD-577. Merging in the current state would publish placeholder text and internal TODOs to the live documentation site.

Test suggestions

• Verify that the 'Auto-configure repository' section is correctly linked via the 'auto-configure-repository' anchor ID.
• Verify the 'How to turn it on' section accurately reflects the final UI labels for the button and dashboard menus.
• Verify the 'What happens during a run' sequence matches the backend agent's execution flow.
• Verify the documentation of the 'paid' status matches the final plan tier decided in OD-577.
• Verify that the data privacy and redaction notes are consistent with existing platform AI policies.

Prompt proposal for missing tests

Consider implementing these tests if applicable:
1. Verify the 'How to turn it on' section accurately reflects the final UI labels for the button and dashboard menus.
2. Verify the documentation of the 'paid' status matches the final plan tier decided in OD-577.

_{TIP Improve review quality by adding custom instructions
TIP How was this review? Give us feedback}

[andrzej-janczak]

@LipeGheno looks cool, just few comments from me and other.

I think we also can add some screen shots. 📸