{"slug": "autoprover-ai-agents-and-formal-methods-for-intent-specs-bugs-analysis", "title": "AutoProver: AI agents and formal methods for intent, specs, bugs analysis", "summary": "AutoProver, a new tool from Certora, uses AI agents and formal methods to automatically infer intent from code, generate specifications, and prove the absence of bugs, turning plain-language documentation into formal verification rules. The tool generates tests and proofs from intent, reports code and design bugs automatically, and improves accuracy with each review cycle, with pricing starting at a pay-as-you-go credits model.", "body_md": "# AutoProver\n\nAutoProver utilises AI agents and formal methods to automatically infer intent from your code, generate specifications, and prove the absence of bugs.\n\n## AutoProver turns docs into Formal Verification\n\nGenerate tests and proofs from your intent, surface code and design bugs automatically, and improve the specification set with every review cycle.\n\n### Generates tests and proofs from your intent\n\nReads your docs, formalizes them into properties, converts those into fuzz tests and formal verification rules. No spec writing experience needed.### Reports bugs automatically\n\nSpecialized agents execute the tests and rules, handle triaging and troubleshooting, and a dedicated bug-hunting agent looks for unaccounted-for design issues. Results land in a structured report: code bugs, design bugs, property status.### Gets more accurate with every run\n\nAccept or reject findings in plain text. AutoProver updates its understanding of your codebase with each iteration.## How AutoProver works\n\nAutoProver orchestrates a set of AI agents to infer intent, generate specifications, run tests, and report results, with your feedback shaping every iteration.\n\nProblem\n\n### Logic bugs go undetected.\n\nThe trickiest errors are logical, not syntactic. Manual code review is slow, expensive, and does not scale.\n\nSolution\n\nAutoProver identifies issues in your code and design that traditional testing misses.\n\nProblem\n\n### AI agents do not guarantee correctness.\n\nModels improve constantly, but they produce non-deterministic results and have no way to verify intent. Code written with one model can be broken by the next.\n\nSolution\n\nAutoProver infers and formalizes intent, so correctness is not left up to chance.\n\nProblem\n\n### Formal verification requires specialists.\n\nWriting specifications is a difficult craft that most development teams do not have in-house.\n\nSolution\n\nAutoProver generates specifications automatically from your existing documentation or a plain-language design doc.\n\n## From intent to proof, then back into feedback\n\nEach run turns plain-language intent into executable verification work, then uses your review to sharpen the next run.\n\n### Infer intent\n\nAutoProver reads your existing code documentation or a design doc you write in plain language. No formal specification experience needed to get started.\n\n### Generate specifications\n\nIntent is converted into formal properties covering individual modules and cross-module interactions, independent of the implementation.\n\n### Produce tests and proofs\n\nSpecifications become formal rules via Certora Prover, covering both specific scenarios and edge cases.\n\n### Run and triage automatically\n\nSpecialized agents execute the tests and rules, handle troubleshooting, and verify that results are sound. A separate bug-hunting agent looks for issues not covered by the design.\n\n### Review results\n\nAutoProver delivers a report in three parts: code bugs, design bugs, and full property status with test and rule results for each.\n\n### Give feedback and iterate\n\nAccept or reject findings in plain text. AutoProver updates based on your input. Open a PR with the generated tests and rules when you are ready.\n\n## Pricing\n\nCredits-based, pay-per-run. Plans add team seats, discounts, and automation access.\n\n#### Plan Includes\n\n1 Seat\n\nTeam members in your organization\n\nPay-as-you-go\n\nRuns billed from your balance at the standard rate\n\nREST API access\n\nGenerate API keys for your custom integrations.\n\nGitHub Action CI integration\n\nOne-click CI integration on the repos.\n\nPriority support\n\nGet faster help\n\n#### Plan Includes\n\n5 Seats\n\nTeam members in your organization\n\n10% discount\n\nEvery run costs 10% less than the standard rate\n\nREST API access\n\nGenerate API keys for your custom integrations.\n\nGitHub Action CI integration\n\nOne-click CI integration on the repos.\n\nPriority support\n\nGet faster help\n\nFrom\n\n#### Plan Includes\n\nUnlimited Seats\n\nTeam members in your organization\n\nHigher Discount\n\nCustom discount - greater than 10%.\n\n2–36 month terms\n\nMonthly committed credits with a fixed minimum term\n\nREST API access\n\nGenerate API keys for your custom integrations.\n\nGitHub Action CI integration\n\nOne-click CI integration on the repos.\n\nPriority support\n\nGet faster help\n\n## As easy to run as a compiler\n\nA compiler checks that your source code translates correctly to machine code. AutoProver checks that your intent written in plain language translates correctly to what your code actually does.\n\nAvailable for Solidity. Rust, C, C++, Java, and Go coming soon.\n\nFAQ", "url": "https://wpnews.pro/news/autoprover-ai-agents-and-formal-methods-for-intent-specs-bugs-analysis", "canonical_source": "https://app.certora.com/", "published_at": "2026-08-13 20:02:24+00:00", "updated_at": "2026-08-13 20:12:05.298391+00:00", "lang": "en", "topics": ["ai-agents", "ai-tools", "ai-research"], "entities": ["AutoProver", "Certora"], "alternates": {"html": "https://wpnews.pro/news/autoprover-ai-agents-and-formal-methods-for-intent-specs-bugs-analysis", "markdown": "https://wpnews.pro/news/autoprover-ai-agents-and-formal-methods-for-intent-specs-bugs-analysis.md", "text": "https://wpnews.pro/news/autoprover-ai-agents-and-formal-methods-for-intent-specs-bugs-analysis.txt", "jsonld": "https://wpnews.pro/news/autoprover-ai-agents-and-formal-methods-for-intent-specs-bugs-analysis.jsonld"}}