Sign InOpen Brain
AI EngineerVideoSource Linked

Your agents lack context: Here's how to fix "You're absolutely right!" — Brandon Waselnuk, Unblocked

In one same-prompt test, a context engine nearly halved agent token use. The design goal is targeted, permission-aware retrieval that resolves conflicts across engineering systems.

AI Engineer · Sep 9, 2026
Open Source Open MarkdownOpen JSON
Source Summary

Unblocked ran the same prompt and model with and without its context engine. The unassisted run used **21 million tokens**; the context-backed run used **10.8 million** and reportedly saved **2 hours** of wall-clock time.

Practical Implication

For larger coding-agent jobs, retrieve organizational context before search and rework compound. Combine semantic retrieval with deterministic queries, rank conflicting sources, personalize by team and repository, and enforce the requester’s permissions.

Agent-Ready Context
Unblocked ran the same prompt and model with and without its context engine. The unassisted run used **21 million tokens**; the context-backed run used **10.8 million** and reportedly saved **2 hours** of wall-clock time.

For larger coding-agent jobs, retrieve organizational context before search and rework compound. Combine semantic retrieval with deterministic queries, rank conflicting sources, personalize by team and repository, and enforce the requester’s permissions.

These figures come from one vendor-presented task, without enough detail here to generalize the reduction. Curated files can become stale, while an MCP connection alone does not ensure the agent calls the right source or continues searching.
Connected Context · Feed7 Judgment

This adds a vendor-reported cost and latency result to the case for retrieving organizational context before coding work sprawls. It confirms that repository access alone is insufficient and sharpens the required context layer—mixed retrieval, conflict ranking, personalization, and permissions—while one opaque comparison cannot establish general savings or retrieval quality.

How to Generate Mergeable Code with a Context Engine — Peter Werry, UnblockedIt supplies a reported efficiency outcome for the same broader retrieval design: surface decisions, conventions, history, and expertise before planning, not merely plausible code.Building Agents Is Trivial Now, Context Is the Next Frontier — Jeff Ng, UnblockedThe earlier outage-trigger example explains why conflict ranking and freshness matter: later organizational evidence can overturn advice that remains technically plausible from repository context.The Last Human Code Review: Building Trust in AI-Generated Code — Itamar Friedman, QodoThe context engine described here is a prerequisite for the inspectable team rules, incidents, architecture, and contracts that the automated-review proposal expects agents to use.tirth8205/code-review-graphThe graph project offers a narrower implementation for deterministic code relationships, complementing the target’s semantic retrieval of broader organizational knowledge rather than replacing it.
Context Map
contextcoding#context-engineering#retrieval#agent-reliability
Uncertainty
These figures come from one vendor-presented task, without enough detail here to generalize the reduction. Curated files can become stale, while an MCP connection alone does not ensure the agent calls the right source or continues searching.