↓ Skip to main content
tomrochette.com
  1. Agents/
  2. Context engines/

Context Engines Feature Matrix

Author
glm-5.3, glm-5.3-flash
History / Edit / PDF / EPUB / MD
Created: / Updated: / Status: finished / Readability: 3 / 10 minutes read (~1808 words)
agent-curated 309 code-retrieval 6 comparison 24 context-engines 9 developer-tools 25 fully-ai-generated 407 llm=glm-5.3 170 llm=glm-5.3-flash 320
Table of Contents

This matrix compares the nine context tools profiled in this section, feature by feature, so the shortlisting step does not require reading nine notes.

The interesting question is not which engine is best but whether a repository needs one at all: most codebases sit below the only published payback threshold in the category, and I claim most buyers of these engines are paying for an index their own vendors’ data cannot justify.

Legend: ✓ supported, ✗ not supported, ~ partial or conditional, ? not verified. Each column links to the full research note; every cell traces to a source cited there or in the references.

The matrix
#

Feature Augment Code Graft Graphify qmd Repomix rtk Semble Serena Sourcegraph code context platform
Kind coding platform local context-graph CLI local knowledge-graph CLI local search engine local CLI CLI output proxy local search index MCP semantic-code toolkit search platform
Deployment cloud SaaS local CLI, MCP, and repo files; Trail Brain is the hosted upsell local CLI, hosted plans or self-host local CLI plus daemon local CLI local binary local CLI, MCP, or library local MCP server (stdio or HTTP), paid JetBrains plugin backend single-tenant cloud or self-host
Open source ~ harness forks OSS Pi ✓ MIT ✓ Apache-2.0 ✓ MIT ✓ MIT ✓ Apache-2.0 ✓ MIT ~ application GPL-3.0-or-later, SolidLSP MIT ~ SCIP only
Free tier ✗ none ✓ entirely ✓ CLI entirely ✓ entirely ✓ entirely ✓ CLI entirely ✓ entirely ✓ core entirely; JetBrains backend paid ~ public search only
Index model real-time semantic index tree-sitter wiring graph plus optional LLM-written markdown nodes, no vectors deterministic AST knowledge graph, no vectors SQLite FTS5 plus vectors, markdown chunks none, whole-repo pack none, per-command output filtering static embeddings plus BM25 fused, tree-sitter chunks none, live language-server parse (40+ languages); optional JetBrains analysis indexed search plus SCIP intel
Index freshness real-time structural re-sync per query, background rebuild after edits snapshot at build time indexed, refreshed on update snapshot at pack time live per command cached index, auto-invalidated on change live parse of the working tree, no index index-dependent
Delivery to agents own harness only instruction files in 9 agents, 6-tool MCP server, Claude Code hooks and statusline skill in 17 assistants (vendor count), MCP, CLI CLI, MCP server, SDK, plugin CLI pack, ~ MCP hooks rewriting commands in 16 tools MCP, CLI, AGENTS.md instructions, sub-agent installer MCP server (stdio or HTTP) MCP server
Scale where it pays large private repos large repos where agents re-explore, no published threshold repo-scale Q&A and path tracing personal docs and knowledge bases under a few hundred K tokens long interactive sessions with noisy commands repos where grep-and-read burns tokens large polyglot repos, reference hunts and refactors 400K+ LOC
Writes code ✓ agents and factory ✗ maps, queries, blast radius ✗ graphs and queries ✗ searches only ✗ packs only ✗ filters output ✗ searches only ✓ symbol-level edits and renames ~ migrations, beta
Pricing model $20/$100 flat tiers plus usage free, MIT; Trail Brain from $20k/yr is the upsell free core, Pro $10/mo yearly or $15 monthly, Teams $20/seat/mo yearly or $29 monthly, Enterprise early access free, MIT free, MIT free CLI, Pro unpriced free, MIT free core; paid JetBrains plugin, price unpublished from $16K/year
Enterprise orientation ✓ SOC 2, ISO 42001 ~ Trail Brain: SOC 2 Type II all plans, HIPAA BAA on Large ~ hosted Teams/Enterprise plans ✗ ✗ ~ Pro tier, on-prem option ✗ ~ paid JetBrains backend, no enterprise plan ✓ SOC 2, ISO 27001

Reading the matrix
#

This is not one market: a platform, a code-map CLI, a graph CLI, a search engine, a packer CLI, an output proxy, an on-demand search index, an LSP symbol toolkit, and a search platform share a category label but sell nine different jobs. Augment sells the author-review-verify loop around its engine; Graft sells a readable map the agent opens like any other file; Graphify sells structural reasoning about one codebase; qmd sells local retrieval over your documents; Repomix sells one deterministic file; rtk sells cheaper command output; Semble sells instant query-time snippets with no standing service; Serena sells the IDE’s own symbol tools to whatever agent you already run; Sourcegraph sells retrieval to whatever agent you already run. The review service that used to sit in this table, Greptile, moved to the Code review category, because what it sells is judgment on the PR stream, not retrieval. The Writes code row makes the split visible: Augment ships authoring agents, Sourcegraph’s Agentic Batch Changes stays in the migration lane, Serena edits at the symbol level when the task asks, and the other 2026 columns refuse the code-writing job entirely.

I read the delivery row as the lock-in axis the marketing never names. Sourcegraph speaks MCP (plus API and CLI surfaces) to any agent; Repomix hands a plain file to anything that reads; Semble installs itself into whatever agents it finds, MCP, AGENTS.md instructions, or a sub-agent; Graft goes furthest, committing hooks, a statusline, and MCP config into the repo so the wiring travels with the code; Augment’s Context Engine only drives Augment’s own surfaces, so its documented token savings are purchasable only inside its own harness.

The scale row is where the budget decision lives, and the only published threshold belongs to Sourcegraph. Its own CodeScaleBench reports a +0.259 reward delta with agents 30% cheaper and 38% faster in the 400K-2M LOC range, and a slightly negative -0.080 below 400K LOC. Repomix inverts that curve: it pays off under a few hundred thousand tokens, then whole-window packing degrades linearly. Augment and Graft claim the large-private-repo end but publish no size threshold.

Freshness splits real-time from snapshot, and it bites exactly when an agent is mid-edit. Augment indexes in real time; Repomix is a snapshot invalidated by every edit; Sourcegraph depends on index lag it inherits from its architecture; Serena sidesteps the axis entirely by parsing the working tree live with no index at all.

Every efficiency number in this matrix is vendor-run, and the pricing floors span free to $150K. Augment’s 33% token savings, Sourcegraph’s cost deltas, Semble’s 99%-fewer-tokens benchmark, and Graft’s 42% token savings and 54%-to-66% SWE-bench jump all come from the vendors themselves; Augment, Semble, and Graft at least publish methodology alongside the numbers, which is the category’s best practice even if no third party has replicated any of it, and Semble’s founders explicitly decline to claim end-to-end agent improvements.

Choosing from the matrix
#

  • Multi-repo organization past 400K LOC with agents thrashing on local search and an enterprise budget: Sourcegraph.
  • Token-heavy team on one large private repo wanting a single vendor for authoring and review: Augment, after re-running its benchmark on your own repo first.
  • AI review of pull requests rather than retrieval: the Code review category, compared in its own feature matrix.
  • Agents re-discovering how one large codebase connects, with structure and citations preferred: Graphify, after verifying its self-published benchmarks on your repo.
  • Agents starting every session blind on a large repo, and a team that wants the map committed as wiring and regenerated per machine: Graft, keeping the free structural layer and re-running its vendor benchmarks on your repo first.
  • Local search over personal docs, notes, and knowledge bases for humans and agents: qmd, it is free and local.
  • Small or mid repo, one-shot whole-repo questions, onboarding packs, or a CI guard on context budget: Repomix, it is free.
  • Repo too big to pack, agents burning tokens on grep-and-read, no appetite for a standing service: Semble, measuring with semble savings before believing the benchmark.
  • IDE-grade navigation, reference hunts, and symbol edits on a large polyglot repo, free and local: Serena, keeping grep for the vague concept queries symbol tools cannot answer.
  • Metered-API sessions dominated by noisy test, git, and search output: rtk, measuring with rtk gain before believing the savings.
  • Code must stay on-device: Graphify, Repomix, and Graft’s structural layer locally, qmd and rtk entirely, or Sourcegraph self-hosted; Augment’s engine stays in its cloud.

Changes
#

  • 2026-08-24 - Created with four columns as one of the remaining categories’ companion matrices.
  • 2026-08-30 - Graphify, qmd, and rtk columns added, matrix at seven columns, kind-row prose and choosing list extended.
  • 2026-08-30 - Extended to eight columns with Semble inserted alphabetically, with reading, choosing, and references sections extended.
  • 2026-08-30 - Greptile column removed, back to seven columns, when the note moved to the Code review category.
  • 2026-09-16 - Re-dated the re-verification and updated the Graphify pricing cell for the new monthly billing options and the early-access Enterprise tier.
  • 2026-09-20 - Repointed the Graft references to the canonical trailhq/Graft repository after the GitHub org rename; no cells moved.
  • 2026-09-24 - Renamed the Sourcegraph column to its listing title, Sourcegraph code context platform; no cells moved.
  • 2026-09-24 - Removed the verification preamble line on owner request.
  • 2026-09-25 - Updated the Graphify delivery cell to the vendor’s documented 17-assistant installer surface.
  • 2026-10-01 - Reworded banned-term words out of the prose; meaning unchanged.
  • 2026-10-04 - Extended from eight to nine columns with Serena (the LSP-backed MCP semantic-code toolkit), inserted in sorted position between Semble and Sourcegraph and traced to the new note; the intro, reading, and choosing sections updated for the ninth job and the live-parse freshness column.

See also
#

References
#