Add agent-cost-guardrails

[sapph1re]

Adds agent-cost-guardrails — framework-native budget limits and circuit breakers for CrewAI, AutoGen, and LangGraph agents. Enforces per-run token caps and cost thresholds without manual monitoring. Alphabetically placed between agentbox and agent-deck.

Summary by CodeRabbit

• Documentation
  • Added the agent-cost-guardrails tool entry.
  • Documented Python budget tracking and circuit-breaker support for CrewAI, AutoGen/AG2, and LangGraph.
  • Clarified pre-call token estimation and post-response token-cost recording.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: 381d5ebc-0b87-4c52-a822-8578f83b87a1

📥 Commits

Reviewing files that changed from the base of the PR and between 079eac5 and bf3beb6.

📒 Files selected for processing (1)

• README.md

🚧 Files skipped from review as they are similar to previous changes (1)

• README.md

---

📝 Walkthrough

Walkthrough

The README adds an agent-cost-guardrails entry. The entry describes framework adapters, pre-call token checks, and post-response cost recording.

Changes

Tool listing

Layer / File(s)	Summary
Update agent-cost-guardrails entry README.md	Describes CrewAI, AutoGen/AG2, and LangGraph adapters, pre-call token estimation, and post-response token-cost recording.

Estimated code review effort: 1 (Trivial) | ~2 minutes

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly and concisely identifies the added agent-cost-guardrails entry.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@README.md`:
- Line 57: Update the agent-cost-guardrails README description to replace
“per-run token caps” with wording that accurately states support for per-call
token limits and rate limits; do not claim cumulative per-run token limits
unless the implementation and documentation support them.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: a3cc5fef-8d30-4a46-933e-0ffcadd04b9b

📥 Commits

Reviewing files that changed from the base of the PR and between 528ee6f and 05afb60.

📒 Files selected for processing (1)

• README.md

[PierrunoYT]

Thanks for the submission. The library qualifies, but “framework-native” overstates the relationship: these are third-party adapters. The cumulative budget is also process-local middleware, not a provider-side spending cap, and pre-call enforcement depends on token estimates.

Suggested replacement:

agent-cost-guardrails - Python budget tracking and circuit-breaker middleware with adapters for CrewAI, AutoGen/AG2, and LangGraph; checks estimated tokens before calls and records actual token cost after responses.

[sapph1re]

Thanks for the detailed feedback! Updated the description to use your suggested wording:

Python budget tracking and circuit-breaker middleware with adapters for CrewAI, AutoGen/AG2, and LangGraph; checks estimated tokens before calls and records actual token cost after responses.

Committed in bf3beb6.

[sapph1re]

Just a gentle ping — the description has been updated in bf3beb6 to use your suggested wording. Happy to make any further adjustments if needed!

[sapph1re]

Friendly bump — the description was updated in bf3beb6 on 2026-08-04 to match your suggested wording exactly. CodeRabbit's latest pass shows no actionable comments. Let me know if there's anything else needed!