Vercel Workflow

Skip to content What this pays sorona.xyz pays 1.25× the program's reward, up to +$1,000 per finding, plus a complexity bonus of up to $1,000 for especially complex or well-chained bugs. Finding │ Program pays │ Through sorona.xyz Critical (Tier 1 asset) │ $10,000 │ $11,000 High │ $5,000 │ $6,000 Medium │ $1,000 │ $1,250 Low │ $500 │ $625 Submit a vulnerability CVE history 1 Last 12 months 1 Lifetime Access control 35 % Auth bypass 25 % Injection 20 % Code execution 15 % Latest advisories Loading the feed Overview Workflow is Vercel's durable-execution primitive for the JavaScript ecosystem, positioned as first-party infrastructure for Next.js applications and shipped as the default answer for long-running or async work on the largest managed frontend platform. It is young at around 2,263 stars but pushed daily, and its trajectory matters more than its current count: a step-injection bug lands inside the orchestration layer of business-critical async jobs such as payments, provisioning, and agent loops, rather than in a leaf library. Vercel itself flags scheduling and state-persistence bugs as the highest blast radius on this asset. It pays as a Tier 1 asset on the open-source program, from $500 for a low finding up to $10,000 for a critical, and the top bounty band actually awarded runs to $10,250, so the critical row is being paid rather than merely advertised. sorona.xyz adds 1.25× on top of whatever Vercel awards, up to $1,000 more, plus a complexity bonus of up to $1,000 for a well-chained finding. Response times 1 business day First response 7 business days Triage 10 business days after triage Bounty In scope The vercel/workflow repository itself, listed under Tier 1 (High Priority) as a source-code asset that is bounty-eligible up to critical severity, with the asset instruction repeating the tiering The workflow runtime: how a step is scheduled, dispatched, retried and resumed after a crash Durable state, meaning the journal the runtime writes, its serialization format, and the replay that reconstructs a run from it Step completion handling, covering acknowledgement, idempotency and whatever proves a completion genuine The HTTP entry points the framework mounts to receive events, callbacks and completions, along with the authentication on them The tenant and workflow-identity boundary that keeps one party's runs, state and triggers away from another's Event handling, plus the hooks, sleeps, timers and webhooks a workflow can suspend on The published npm package and its supply chain, which the policy names in scope Out of scope Vulnerabilities that arise from a developer's own workflow design, such as a step they wrote with no authorization check Static-analysis or scanner output submitted without a working exploit chain Unverified findings produced by AI or LLM tooling Dependency CVE reports without a demonstrated chain through Workflow in context Issues that only reproduce in a canary or release-candidate build already superseded by a stable release Issues that require the attacker to already hold privileged access to the workflow or its host Cloud-provider and deployment-specific issues, which belong to Vercel's main platform program rather than the open-source one Reports without the mandatory working proof-of-concept zip demonstrating the chain on a production-equivalent deployment Most-valued findings Injection into step scheduling, state persistence, or event handling, which Vercel names as the highest-blast-radius class on this asset and the route to the $10,000 critical row Step injection that makes the runtime execute a step absent from the original workflow definition Cross-workflow access or tenant-context escape, where one party reads state from or triggers steps in another party's workflow Step bypass that skips a step performing an authorization or validation check, which is an auth bypass wearing orchestration clothes Replay or forgery of step completions, letting an attacker mark work as done that the runtime never actually performed Program rules Submit through sorona.xyz first. We validate, coordinate with the program's security team, and add the payout uplift on top of whatever it awards. Include a working proof-of-concept zip that demonstrates the complete exploit chain on a production-equivalent deployment SAST output, scanner output, and unverified AI or LLM findings are ineligible The bug has to be in the framework itself, not in a developer's own workflow design Issues that require the attacker to already hold privileged access are out of scope, as are canary or release-candidate bugs already superseded by a stable release Vercel's confidentiality obligations run for two years and survive the end of the program Program stats 100 Reports resolved $241k Bounties paid 89% Response efficiency Disclosure Coordinated through Vercel. Report contents stay confidential until Vercel consents in writing to their release, so there is no fixed publication clock. sorona.xyz holds your finding until Vercel has resolved it. Safe harbor Sorona will not pursue legal action against you for good-faith research submitted to us in line with this policy. We cannot speak for the upstream project, its vendor, or any third party, and we do not grant authorization on their behalf. You are responsible for understanding and following the laws that apply to you, wherever you are, and for staying within any scope the project has defined. Submit a vulnerability