[Mechanism] Update Bounties page with agent-compatible bounty pattern

[vinaykrsinghal-stage]

Slug

bounties

Short Description

Update the Bounties mechanism page with an emerging agent-compatible bounty pattern.

Tags

task-based, execution, open, ai-agents

Featured

No.

Banner Image

Existing page edit: keep current banner at /content-images/mechanisms/bounties/banner.png.

Description

This is a targeted update to the existing bounties mechanism page, not a new mechanism page.

Target page: https://gitcoin.co/mechanisms/bounties

Source file: https://github.com/gitcoinco/gitcoin_co_30/blob/main/src/content/mechanisms/bounties.md

Rationale

Gitcoin already has research on AI agents and public goods, including agent discovery, evaluation, monitoring, and the need for human oversight:

https://gitcoin.co/research/ai-agents-and-public-goods-the-emerging-agentic-economy

The current Bounties page explains task-based funding but does not mention agent-compatible bounty design. This edit adds a concise, practical note about when bounties are suitable for AI agents and what extra structure funders need.

Add to ## Limitations

- Agent-compatible bounties need extra structure: explicit eligibility, deadlines, submission formats, prohibited behavior, and payment-claim rules so agents and human operators do not have to infer what is allowed.

Add under ## Examples and Use Cases, after ### Design and Graphics

### Agent-Compatible Tasks
As AI agents become capable of producing reviewable artifacts, bounties can be structured so agents handle bounded execution while humans retain accountability. Good fits include documentation updates, passive code review, public data analysis, ecosystem maps, and research briefs. These tasks work best when eligibility, deadlines, deliverable formats, and human payout or claim steps are explicit.

Add to ## Further Reading

- [Gitcoin — AI Agents and Public Goods: The Emerging Agentic Economy](https://gitcoin.co/research/ai-agents-and-public-goods-the-emerging-agentic-economy)

Patch

diff --git a/src/content/mechanisms/bounties.md b/src/content/mechanisms/bounties.md
index 0000000..0000000 100644
--- a/src/content/mechanisms/bounties.md
+++ b/src/content/mechanisms/bounties.md
@@ -36,6 +36,7 @@ Bounties create a direct link between defined work and compensation.
 - Not suited for long-term relationship building
 - Poorly handles ambiguous or evolving requirements
 - Requires clear scope — if the task can't be well-defined, bounties won't work
+- Agent-compatible bounties need extra structure: explicit eligibility, deadlines, submission formats, prohibited behavior, and payment-claim rules so agents and human operators do not have to infer what is allowed.
 
 ## Best Used When
 
@@ -56,6 +57,10 @@ Community labs fund multi-language resource translations via bounties, allowing
 ### Design and Graphics
 DAOs issue bounties for protocol explanatory graphics, social media assets, and UI/UX improvements.
 
+### Agent-Compatible Tasks
+As AI agents become capable of producing reviewable artifacts, bounties can be structured so agents handle bounded execution while humans retain accountability. Good fits include documentation updates, passive code review, public data analysis, ecosystem maps, and research briefs. These tasks work best when eligibility, deadlines, deliverable formats, and human payout or claim steps are explicit.
+
 ## Further Reading
 
 - [Allo Capital — Bounties](https://www.allo.capital/mechanisms/bounties)
+- [Gitcoin — AI Agents and Public Goods: The Emerging Agentic Economy](https://gitcoin.co/research/ai-agents-and-public-goods-the-emerging-agentic-economy)

Authors

vinaykrsinghal-stage | https://github.com/vinaykrsinghal-stage

Related Apps

gitcoin-grants-stack, allo-protocol

Related Mechanisms

bounties, direct-grants, requests-for-proposals

Related Case Studies

N/A

Related Research

ai-agents-and-public-goods-the-emerging-agentic-economy

Related Campaigns

N/A

Submission Checklist

• My short description is accurate and fits in one line
• My description is complete and written in Markdown