BountyVerdict Agent Decision Tools
About
Six read-only decision tools for coding agents to assess GitHub bounties, audit agent instructions, diagnose GitHub Actions failures and flakes, and detect MCP tool drift. Remote Streamable HTTP
Details
- Author
- cristianmoroaica
- Categories
- Developer Tools, Other
Jump to
Setup
Install BountyVerdict Agent Decision Tools in your MCP client (Claude Desktop, Cursor, Windsurf, and others).
Repository: https://github.com/cristianmoroaica/bountyverdict
Follow the installation instructions in the repository README, then restart your MCP client.
Six read-only decision tools for coding agents to assess GitHub bounties, audit agent instructions, diagnose GitHub Actions failures and flakes, and detect MCP tool drift. Remote Streamable HTTP
Seven paid, bounded decision APIs for autonomous coding agents. Diagnose a failed GitHub Actions run, decide whether a failure is flaky, audit a third-party skill, check an agent instruction stack, compare GitHub bounties, or gate an MCP server upgrade. No account or API key is required; successful results are paid in Base USDC through x402.
Install the router that selects the narrowest check and applies its payment safeguards:
npx skills add cristianmoroaica/bountyverdict --skill route-github-agent-checks -y
Or inspect and install the immutable v1.0.3 router with GitHub's native skill workflow:
gh skill preview cristianmoroaica/bountyverdict route-github-agent-checks@v1.0.3 gh skill install cristianmoroaica/bountyverdict route-github-agent-checks --pin v1.0.3
GitHub Copilot CLI can install the repository as a plugin without a separate marketplace setup:
copilot plugin install cristianmoroaica/bountyverdict
That plugin exposes the five task-specific public GitHub, CI, and MCP workflow skills. SkillVerdict and the broad router remain separate least-privilege installs.
- “Why did this public GitHub Actions workflow fail, and what should I do next?”
- “Will this MCPtools/listschema change break my agent after the server upgrade?”
Task-specific skills are the least-privilege path. Install all seven only when needed withnpx skills add cristianmoroaica/bountyverdict --skill '*' -y.
For the six-tool remote MCP surface, install the task-first marketplace adapter directly from its public repository while directory indexing is pending:
npx skills add cristianmoroaica/bountyverdict-mcp-skill --skill route-github-agent-decisions
That adapter routes GitHub bounty, repository-readiness, GitHub Actions, flaky-retry, and MCP schema-drift requests to one exact tool, keeps SkillVerdict excluded, and uses a coarse source marker without storing prompts, tool arguments, or caller identities. A valid unsigned call is only a free payment preview; an x402-aware client must separately authorize the exact Base USDC amount before receiving a verdict.
MCP-compatible agents can instead connect to the production Streamable HTTP server athttps://bountyverdict-agent-production.mimirslab.workers.dev/mcp. It exposes six real paid tools matching the independently distributed products; SkillVerdict remains excluded from this channel.tools/listpublishes task-selection boundaries, canonical input patterns, and compact machine-readable success contracts covering each verdict or action and its reuse rule. Invalid input is rejected before payment, while a valid selected tool returns an exact x402 USDC requirement. The official registry contract isserver.jsonunderio.github.cristianmoroaica/bountyverdict; its publisher metadata includes unbranded task keywords, use cases, prices, and result summaries so downstream MCP aggregators can index what each tool actually solves.
Agentic Resource Discovery crawlers can ingest the origin-ownedai-catalog.json. It advertises the existing MCP server with six unbranded representative buyer queries for semantic retrieval; catalog fetches are measured separately from tool calls, payments, and purchases.
For the broad CI use case, the crawlableGitHub Actions Failure Diagnosis MCP Serverguide compares the root-cause and flaky-retry tools, leads with the source-marked routing-adapter install, and publishes exact prices, free samples, typed outputs, and the mutation boundary.
Every product has a free sample, a machine-readable OpenAPI contract, a declared price, and a successful-resultservice_reuserule. Invalid inputs and upstream failures are not settled. Start with theagent page,agent-manifest.json, the productionopenapi.json, or the remote MCP server above.
The bounty samples are captured from real public issues rather than invented fixtures: thesingle-check samplerejects an assigned, withdrawn, and maintainer-rejected TypeORM bounty, while theportfolio samplerefuses to recommend either it or an open but already-assigned Tenstorrent$1,500bounty. Use the$0.05single check for one candidate; the fixed$0.40portfolio is economical for 8–10 candidates or when one ranked, partial-failure-aware response materially simplifies the workflow.
The seven contracts are continuously checked in production. Coinbase Bazaar currently merchant-indexes six resources. Six existing non-SkillVerdict products are also available through exact-schema the402 listings, either per call or through a $1.05 monthly bundle of up to 20 combined requests. Owner-funded proofs are excluded from customer revenue.
For the strongest current buyer-intent fit, inspect thefree SkillVerdict sampleand the task-specificpreflight-agent-skillsworkflow. It audits the entire commit-pinned public skill directory without executing it; the paid verdict costs $0.06.
HarnessVerdict pins the repository default branch to an immutable commit and audits recognizedAGENTS.md,CLAUDE.md,GEMINI.md, Copilot, Cursor, andSKILL.mdsurfaces without cloning or executing repository code. It reports evidence-linked path, scope, portability, context-budget, skill-frontmatter, and secret-like-material findings.
SkillVerdict pins and statically scans a requested public skill directory without executing it. It combines high-confidence dangerous-pattern checks with repository context, capability disclosure, external-domain inventory, secret redaction, and explicit coverage to reduce both missed supply-chain hazards and naive false positives.
RunVerdict reads exact-attempt job metadata and bounded failed-job logs without executing or rerunning code. It separates primary failures from aggregate-result jobs, redacts secret-like excerpts, classifies root-cause families, and recommends whether to fix, investigate, wait, or retry.
FlakeVerdict compares a completed public GitHub Actions failure with other attempts of the same run, same-commit outcomes, and up to 12 earlier comparable workflow runs. It scans at most 8 selected failed-job logs and 4 MiB of log data, never executes repository code, and never triggers, reruns, cancels, approves, or otherwise mutates CI. Its six typed outcomes areCONFIRMED_FLAKE,LIKELY_FLAKE,RECURRING_FAILURE,NEW_FAILURE,INCONCLUSIVE, andNOT_FAILED; only a currentCONFIRMED_FLAKEcan recommend one retry.
Agents can inspect thefree FlakeVerdict sampleand use the guardedclassify-github-flakes skill. Every successful result carries this reuse rule: call FlakeVerdict for every completed public GitHub Actions failure before spending a retry; reuse the result only for its exact run ID and attempt, and call again after a new attempt appears.
MCPDriftVerdict accepts two complete inline MCP 2025-11-25tools/listsnapshots, canonicalizes and hashes them, proves only a conservative JSON Schema 2020-12 compatibility subset, reverses variance for outputs, and flags tool removals, model-facing metadata changes, task-mode breaks, and declared safety-hint regressions. It validates and computes the whole verdict before x402 settlement and never connects to an MCP server, fetches catalog URLs, invokes tools, or follows catalog instructions. Agents can inspect thefree sampleand guardedcheck-mcp-tool-drift skill.
Seeagent/README.mdfor the protocol, local verification, and deployment configuration.
The launch prices and differentiation are grounded in a live Bazaar comparison documented indocs/MARKET_VALIDATION.md. Agents and crawlers can readllms.txtbefore deciding whether a product is relevant.
Review the publicsecurity policyandprivacy/data-handling disclosurebefore submitting data or authorizing payment. Vulnerabilities can be reported privately through GitHub without opening a public issue.
The score is deliberately conservative and deterministic. BountyVerdict currently reads up to 300 issue comments and the first and newest timeline pages. Very large threads may contain additional evidence it does not see. Anonymous GitHub API rate limits apply.
Treat aVIABLEverdict as permission to investigate further—not permission to start coding. Reproduce the issue, read contribution and AI-use policies, confirm reward terms and payout eligibility, and establish acceptance criteria first.
This is a web browser that enables your coding agent, such as Claude Code, to visit websites on your behalf and assist you in identifying bugs or creating UI test cases.
Help agents automatically write and test stories for your UI components
AI-Safe Code Analysis with 113+ MCP tools for guard validation, memory, workflow, and testing.
Integrate with CodeRabbit AI for automated code reviews, pull request analysis, and report generation.
fable-discipline is a Claude Code plugin that makes agentic software work follow repeatable working patterns: design before code, verify after edits, separate author from reviewer, preserve verified state between sessions, and report uncertainty honestly.
MCP server that generates production-grade engineering standards (SOLID, testing, architecture, CI/CD) for AI coding assistants
An intelligent MCP server that automates the reproduction of GitHub issues for AWS CDK projects.
A Grey Hack server for Cursor IDE, providing GitHub code search, Greybel-JS transpilation, API validation, and script generation.
Execute JavaScript code in a modern runtime environment with support for various built-in modules.
Provides a secure JavaScript execution environment for running code snippets.
A server for JavaScript/TypeScript development with intelligent project tooling and testing capabilities.
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.
