The context layer for modern engineering teams Unblocked turns code, docs, tickets, and conversations into actionable context, so engineers move faster and agents stay on track. Get Started for Free Book a Demo Trusted by leading engineering teams Understanding how your system works takes time You have to piece together code, conversations, tickets, and docs to understand what’s current, what’s trusted, and what actually happens in production. Notion 6mo ago System Architecture Tokens validated by Auth Service. Outdated docs GitHub 14d ago Refactor auth flow — #3243 Replaced legacy JWT flow. Outdated change Datadog 4 hrs ago auth.latency.p95 Spikes correlated with deploys. Symptom, not cause Slack 3d ago #eng-discussion Internal traffic skips the check. Tribal knowledge Jira 5d ago UNB-1337 Rejecting valid sessions. Unknown status Confluence 1mo ago Auth-Service Overview All requests check tokens. Conflicting info How does auth validation work? Agents inherit the same problem They retrieve what’s easy to find, miss what matters, and waste time and tokens generating from incomplete context. unblocked TS token-service.ts 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 import jwt from 'jsonwebtoken' import { logger } from '../lib/logger' const JWT_SECRET = process . env . JWT_SECRET ! const ISSUER = 'auth.unblocked.io' export interface TokenPayload { sub : string org : string scope : string [] } export async function validateToken ( token : string , ): Promise < TokenPayload > { const payload = jwt . verify ( token , JWT_SECRET , { issuer : ISSUER , algorithms : [ 'RS256' ], }) as TokenPayload Agent Chat How does auth validation work? Searching for context... github_mcp notion_mcp slack_mcp datadog_mcp jira_mcp confluence_mcp Notion 6mo ago System Architecture Tokens validated by Auth Service. GitHub 14d ago Refactor auth flow — #3243 Replaced legacy JWT flow. Datadog 4 hrs ago auth.latency.p95 Spikes correlated with deploys. Slack 3d ago #eng-discussion Internal traffic skips the check. Jira 5d ago UNB-1337 Rejecting valid sessions. Confluence 1mo ago Auth-Service Overview All requests check tokens. Give engineers and agents context they can trust Unblocked reconstructs fragmented context into grounded answers engineers and agents can rely on. unblocked TS token-service.ts 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 import jwt from 'jsonwebtoken' import { logger } from '../lib/logger' const JWT_SECRET = process . env . JWT_SECRET ! const ISSUER = 'auth.unblocked.io' export interface TokenPayload { sub : string org : string scope : string [] } export async function validateToken ( token : string , ): Promise < TokenPayload > { const payload = jwt . verify ( token , JWT_SECRET , { issuer : ISSUER , algorithms : [ 'RS256' ], }) as TokenPayload Agent Chat How does auth validation work? Searching for context... unblocked_context_engine Agent Auth validation runs through token-service.ts (PR #3243). Each request hits the token service first — the JWT signature is checked, the revocation list in Redis is consulted, and a verified principal is stamped onto the request before any downstream service sees it. The Confluence auth spec marks this as canonical and Slack confirms the legacy gateway path is deprecated. One gap remains in UNB-1337: internal service-to-service calls still bypass the check, which matches the Datadog latency anomaly on direct routes. Agent Chat How does auth validation work? Searching for context... unblocked_context_engine Agent Auth validation runs through token-service.ts (PR #3243). Each request hits the token service first — the JWT signature is checked, the revocation list in Redis is consulted, and a verified principal is stamped onto the request before any downstream service sees it. The Confluence auth spec marks this as canonical and Slack confirms the legacy gateway path is deprecated. One gap remains in UNB-1337: internal service-to-service calls still bypass the check, which matches the Datadog latency anomaly on direct routes. UNIFYING CONTEXT RETRIEVING RELEVANT DOCS PERSONALIZING RELEVANCE ENFORCING PERMISSIONS RESOLVING CONFLICTS OPTIMIZING TOKENS Notion 6mo ago System Architecture Tokens validated by Auth Service. GitHub 14d ago Refactor auth flow — #3243 Replaced legacy JWT flow. Datadog 4 hrs ago auth.latency.p95 Spikes correlated with deploys. Slack 3d ago #eng-discussion Internal traffic skips the check. Jira 5d ago UNB-1337 Rejecting valid sessions. Confluence 1mo ago Auth-Service Overview All requests check tokens. 3,500 tokens used unblocked TS token-service.ts 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 import jwt from 'jsonwebtoken' import { logger } from '../lib/logger' const JWT_SECRET = process . env . JWT_SECRET ! const ISSUER = 'auth.unblocked.io' export interface TokenPayload { sub : string org : string scope : string [] } export async function validateToken ( token : string , ): Promise < TokenPayload > { const payload = jwt . verify ( token , JWT_SECRET , { issuer : ISSUER , algorithms : [ 'RS256' ], }) as TokenPayload Agent Chat How does auth validation work? Searching for context... unblocked_context_engine Agent Auth validation runs through token-service.ts (PR #3243). Each request hits the token service first — the JWT signature is checked, the revocation list in Redis is consulted, and a verified principal is stamped onto the request before any downstream service sees it. The Confluence auth spec marks this as canonical and Slack confirms the legacy gateway path is deprecated. One gap remains in UNB-1337: internal service-to-service calls still bypass the check, which matches the Datadog latency anomaly on direct routes. Agent Chat How does auth validation work? Searching for context... unblocked_context_engine Agent Auth validation runs through token-service.ts (PR #3243). Each request hits the token service first — the JWT signature is checked, the revocation list in Redis is consulted, and a verified principal is stamped onto the request before any downstream service sees it. The Confluence auth spec marks this as canonical and Slack confirms the legacy gateway path is deprecated. One gap remains in UNB-1337: internal service-to-service calls still bypass the check, which matches the Datadog latency anomaly on direct routes. UNIFYING CONTEXT RETRIEVING RELEVANT DOCS PERSONALIZING RELEVANCE ENFORCING PERMISSIONS RESOLVING CONFLICTS OPTIMIZING TOKENS Notion 6mo ago System Architecture Tokens validated by Auth Service. GitHub 14d ago Refactor auth flow — #3243 Replaced legacy JWT flow. Datadog 4 hrs ago auth.latency.p95 Spikes correlated with deploys. Slack 3d ago #eng-discussion Internal traffic skips the check. Jira 5d ago UNB-1337 Rejecting valid sessions. Confluence 1mo ago Auth-Service Overview All requests check tokens. 3,500 tokens used How the context engine works Unblocked builds a model of your engineering system from your code, docs, tickets, and conversations. We manage the hard parts — freshness, source deconfliction, permission awareness, and personalization — so you and your agents get fast, reliable answers and direction. Connect your sources Unblocked indexes your codebase, documentation, tickets, and conversations — building a unified model of how your system works and why it works that way. Build living context Unblocked understands relationships between code, decisions, and conversations. It tracks conventions, past choices, and active work — and stays current as your codebase evolves. Put it to work Surface context wherever your team works — in Slack, the terminal, the web, or via MCP directly into your AI agents. Engineers get instant answers to whatever’s blocking them. Agents get the context they need to stay on track. How teams use Unblocked Every engineering workflow depends on context that's buried in code, docs, tickets, and conversations. Unblocked automatically surfaces that context so you don't have to dig and agents don't have to guess. Get answers instantly Ask anything about your codebase, architecture, or past decisions. Draws from code, PRs, Slack, and docs with sources. Plan with full context Before your agent starts a task, give it the context to stay on track: patterns, decisions, and prior conversations. Review code in context Review in the context of team conventions and prior art — not just what it does, but whether it fits. Onboard without slowing the team Every new hire gets fast answers without interrupting the rest of the team. Assemble incident data in seconds Commits, config changes, and Slack activity — without opening 12 tabs. Investigate bugs without any digging Quickly identify whether a bug is real, related to recent changes, or has happened before. Unblocked works everywhere you do Connect Unblocked to your tools and agents in minutes. Use MCP, CLI, API, or direct integrations to bring context into every workflow. MCP Connect Cursor, Claude Code, Copilot and more. CLI Work directly from your terminal. Slack Add @Unblocked to any channel. Web Search, explore, and get answers from a browser. API Build context directly into any API-compatible tool. Built securely for professional teams Unblocked is designed with security and privacy as core requirements. Customer data is isolated, authorized, and encrypted and your data will never be used to train models. Learn more about security SOC 2 TYPE II DATA ISOLATION SSO & SCIM SUPPORT ENCRYPTION IN TRANSIT & AT REST See what happens when your agents have context “Unblocked acts as a safety blanket against unknown unknowns. The stuff buried in Slack threads you’d never think to look for, but that will bite you if you miss it.” Gustavo Alvarez Senior AI Engineer “Unblocked eliminated work that was genuinely mind numbingly arduous.” Austin Rojan Onboarding Specialist “I’d rate Unblocked’s impact as eight or nine out of ten, and it’s only gotten better.” Justin McCraw Senior Software Engineer