EKOS is live · pricing not published · last verified 26 Oct 2026 · viability monitor
EKOS
EKOS compiles SQL, Pentaho ETL and git history into a cited MCP knowledge ledger your agents can query
EKOS publishes its misses beside its wins: 49.7% correctness, 51.6% groundedness, and the one grep query it loses. Pair that with an evidence ledger where answers cite a specific fragment and you get something worth running on legacy ETL archaeology. Just don't wire it into an unattended production path yet; the numbers are moving, not settled.
Verdict generatedVerified 18d ago · liveness 67/100 · cite: rightaichoice.com/tools/ekos
- Data and platform engineers inheriting Pentaho, SQL or stored-procedure logic with no living documentation
- Teams standardizing on MCP that need agent answers citing exact source fragments, not paraphrase
- Migration and audit projects needing a reproducible, diffable artifact of what legacy ETL actually does
- Organizations able to self-host a compiler pipeline against their own repositories
- Non-technical users wanting a no-code AI data assistant
- Teams that need a fully benchmarked component in an automated production path today
- Anyone without a repository or legacy system to point the compiler at
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip EKOS if you need a no-code assistant or a product with production-grade accuracy today — its own 101-scenario suite scores 49.7% answer correctness, so you must be willing to run, verify and self-host the compiler against your own repository.
You own the compute: ingesting a 2,022-file repository and compiling it to a ledger runs on your machines, so the real cost is cluster time plus the engineer-hours to operate the pipeline.
The seed flags EKOS as freemium and the vendor's own published benchmarks describe a self-hosted pipeline, so your real cost line is engineering time and compute rather than a seat fee. Compare it against evidence-inference and code-search tools that bill per seat or per query: EKOS gives you the ledger locally, but you pay for it in setup and operation. It makes most sense for a team that already has a platform engineer and a repository to point it at, not for a solo buyer looking for a
In short
EKOS — EKOS compiles SQL, Pentaho ETL and git history into a cited MCP knowledge ledger your agents can query. Best for Data and platform engineers inheriting Pentaho, SQL or stored-procedure logic with no living documentation, Teams standardizing on MCP that need agent answers citing exact source fragments, not paraphrase, Migration and audit projects needing a reproducible, diffable artifact of what legacy ETL actually does. Pricing not published.
What is EKOS?
EKOS — EKOS compiles SQL, Pentaho ETL and git history into a cited MCP knowledge ledger your agents can query. Best for Data and platform engineers inheriting Pentaho, SQL or stored-procedure logic with no living documentation, Teams standardizing on MCP that need agent answers citing exact source fragments, not paraphrase, Migration and audit projects needing a reproducible, diffable artifact of what legacy ETL actually does. Pricing not published.
EKOS is an open-source knowledge layer built for code rather than documents. Point its CLI at an enterprise repository and it compiles SQL DDL, Pentaho jobs, git history, Confluence pages and GitHub issues into one Transformation IR, so a Pentaho step and a SQL SELECT resolve to the same node types: Source, Filter, Join, Aggregate, Sink. That typed intermediate representation comes first and retrieval second, which is the line between EKOS and RAG-over-repo assistants. The output is an append-only, content-addressable evidence ledger served read-only over MCP. Agents get answers that name the fragment they drew on instead of paraphrasing it, and the ledger covers issues, PRs and files. Recently the team ran RFC 0056/0057/0058 live end to end on a real open-source repo: compiled metadata plus two real parser bugs found and fixed, a local-LLM NL-to-SQL pipeline executed against a real ClickHouse server, and a multi-step diagnostic question answered over the actual MCP protocol. The workflow is a pipeline, not a chat box: ekos build, recover, resolve, compile, commit, then ekos mcp serve --workspace . to expose the workspace over stdio JSON-RPC to any MCP client. Live query stays an audited exception, with the ClickHouse connector as the worked example. It is aimed at data and platform engineers who inherit Pentaho jobs, stored procedures and enterprise SQL with no living documentation, and at teams already on MCP who want cited answers over production code. EKOS grades itself in public. A 101-scenario, seven-category suite scored by six deterministic evaluators lands at 49.7% answer correctness (up from 42.5%) and 51.6% evidence groundedness. On the first published benchmark, the compiled ledger used 67-93% fewer tokens than raw grep on the same repo and tokenizer, and the single query where grep wins is printed on the page. Against generic repo-chat assistants, the difference is a diffable artifact you own rather than a chat transcript.
Key features
- Compiles SQL DDL, Pentaho jobs, git history, Confluence and GitHub issues into a shared Transformation IR
- Recovers legacy ETL and stored-procedure logic as Source, Filter, Join, Aggregate and Sink nodes
- Serves compiled knowledge read-only over MCP via stdio JSON-RPC to any MCP client
- Answers cite the exact evidence fragment they came from instead of paraphrasing
- Append-only, content-addressable evidence ledger spanning issues, PRs and files
- CLI pipeline: ekos build, recover, resolve, compile, commit, then mcp serve --workspace
- Answers multi-step diagnostic questions by chaining live MCP calls over the real protocol
- Local-LLM NL-to-SQL pipeline run live against a ClickHouse server
- 67-93% fewer tokens than raw grep on the same repo and tokenizer, losing case published
- 101-scenario answer-quality suite across seven categories, graded by six deterministic evaluators
- Cold, timed interpretation of a 2,022-file repository
- Live query gated behind an audited exception; read-only MCP exposure by default
- Setup wizard for Linux, Windows 11 and macOS
- Self-hosted pipeline spanning build, recovery, compilation, storage and search
Behind the verdict
We'd reach for EKOS at the moment a Pentaho job or a 900-line stored procedure outlives the person who wrote it. The pitch is narrow and honest: compile the code into a typed graph, commit the evidence, then let an MCP client ask questions that come back with a cited fragment instead of a confident paraphrase. When to pick it. You have a real repository with legacy ETL, SQL DDL or stored-procedure logic, and you can self-host the pipeline. You already run an MCP client such as Claude or Cursor and want grounded answers over production code. You need an artifact you can diff and review, not a chat log someone has to trust. When to pass. If you want a no-code AI data assistant, this is the wrong shape entirely — the CLI pipeline is the product. If you need a fully benchmarked component in an automated production path, the published suite (49.7% correctness, 51.6% groundedness) says wait. And if nothing in your stack is a legacy system worth compiling, there is nothing to point it at. The closest alternative is repo-chat tooling that embeds files and retrieves chunks. EKOS differs by resolving a Pentaho step and a SQL SELECT to the same node types before retrieval happens. In practice that shows up as token savings: 67-93% fewer than raw grep on the same repo and tokenizer on the first benchmark, with the losing case published alongside. Where it bites. That benchmark is the first number, not a settled one. The token count is a cost signal, not an accuracy guarantee. The 2,022-file cold run surfaced three gaps that were fixed the same day as RFC 0059/0060/0061 — encouraging, but it also means the surface is still being discovered. Live query is gated behind an audited exception for a reason. Treat the read-only ledger as the default and the ClickHouse connector as
Models under the hood
as of 2026-09-09
Limitations
- The vendor's own graded suite reports 49.7% answer correctness and 51.6% evidence groundedness — up from 42.5% correctness on August 3, so it is improving, but roughly half of tested scenarios do not come back right yet.
- Treat output as something to verify rather than trust.
- The tool is CLI- and self-host driven, so you own setup and operation.
- Development is RFC-numbered and moves fast; three new gaps (RFC 0059/0060/0061) were found and fixed the same day during one cold run over a 2,022-file repo.
- Automatic recovery may not cover every custom API, and the vendor explicitly treats unmapped fragments as a normal outcome rather than an error, which means some business logic will land outside the ledger until analyzers catch up.
- Live query is deliberately narrow: the ClickHouse connector (RFC 0056) is an audited exception, and the ledger is otherwise served read-only.
as of 2026-10-09
Researching EKOS? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Is EKOS still active?
EKOS is live · pricing not published · last verified 26 Oct 2026 · viability monitor.
Viability score
How well maintained and how widely used is EKOS? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: November 2026
How we score →What's new in EKOS
Checked 7 days agoAcross the latest 3 updates: 3 news mentions.
The first benchmark number: 67-93% fewer tokens than grep
Raw grep versus the compiled ledger, same real repo and same standard tokenizer, reports 67-93% fewer tokens for the realistic case, with the one case grep wins included rather than hidden.
Interpretation of the whole 2,022-file repo, cold and timed
A cold pipeline run over the entire repo with a graded ekos ask plus MCP Q&A set surfaced three new gaps, reported honestly and fixed the same day as RFC 0059/0060/0061.
Answer quality, measured: 49.7% correctness, 51.6% groundedness
A 101-scenario suite across seven categories graded by six deterministic evaluators, run against the real compiled ledger — correctness up from 42.5%.
What do users say about EKOS?
What people actually say about EKOS — is it worth it?
Our community sentiment research written Aug 13, 2026 is more than 90 days old, so it is being re-reviewed against current data and is not shown.
How we verified this
Every listed fact about EKOS — pricing, features, links, status — is re-read from the vendor's own pages by our verification pipeline and each change is logged with a date; the last pass ran on 26 Oct 2026. Analysis is drafted with AI models and checked against our editorial standards before publishing; see how we verify.
Verification history
We have re-verified EKOS 12 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — changed features, description, who it suits, our verdict
- — changed description, summary, features, our verdict
- — changed our verdict, description, features, summary
- — changed description, summary, our verdict, features
- — changed description, who should skip it, our verdict, summary, features, who it suits, integrations
- — changed summary, integrations, who should skip it, features, who it suits, our verdict, description
Showing the 6 most recent of 12 verification passes.
Free to cite with attribution — this page re-verifies continuously.
How much does EKOS cost in 2026?
EKOS: Pricing not published. Pricing last verified 26 Oct 2026.
Where the pricing makes sense
The company stage and team size where EKOS's pricing actually pencils out — and where peers do it cheaper.
The seed flags EKOS as freemium and the vendor's own published benchmarks describe a self-hosted pipeline, so your real cost line is engineering time and compute rather than a seat fee. Compare it against evidence-inference and code-search tools that bill per seat or per query: EKOS gives you the ledger locally, but you pay for it in setup and operation. It makes most sense for a team that already has a platform engineer and a repository to point it at, not for a solo buyer looking for a
Has EKOS' pricing changed?
No plan price for EKOS has moved since we began logging in Aug 2026; the pricing page is re-read on every verification pass.
Who should pick EKOS?
Pick EKOS if you are Data and platform engineers inheriting Pentaho, SQL or stored-procedure logic with no living documentation, Teams standardizing on MCP that need agent answers citing exact source fragments, not paraphrase, Migration and audit projects needing a reproducible, diffable artifact of what legacy ETL actually does, Organizations able to self-host a compiler pipeline against their own repositories.
Look elsewhere if you are Non-technical users wanting a no-code AI data assistant, Teams that need a fully benchmarked component in an automated production path today, Anyone without a repository or legacy system to point the compiler at, Teams needing an agent that writes back to production systems rather than a read-only ledger.
Skip it if Skip EKOS if you need a no-code assistant or a product with production-grade accuracy today — its own 101-scenario suite scores 49.7% answer correctness, so you must be willing to run, verify and self-host the compiler against your own repository.
GeneratedReal-world workflow fit
Concrete scenarios for the personas EKOS actually fits — and what changes day-one when you adopt it.
You point EKOS at a repo of Pentaho job XML and SQL DDL, run `ekos build && ekos recover && ekos resolve && ekos compile && ekos commit`, and read the recovered Transformation IR as Source, Filter, Join, Aggregate and Sink nodes instead of opening each .kjb by hand.
Outcome: You get a written map of what the pipeline actually does, with each node traceable to the fragment it came from, before you touch the migration.
You start `ekos mcp serve --workspace .` and connect Claude or Cursor over stdio, then ask a multi-step diagnostic question — why a metric spiked — and let the agent chain live MCP calls against the compiled ledger.
Outcome: The agent answers by citing specific fragments rather than pasting raw files, and your token spend on the same question drops 67-93% versus grepping the repo.
You point a local Ollama model at a real ClickHouse server through the audited live-query connector (RFC 0056) for schemas that need current data rather than a compiled snapshot.
Outcome: You get natural-language querying over ClickHouse with the data staying inside your own perimeter and the read-only ledger unchanged for everything else.
Use cases
- Recover and document what a legacy Pentaho or SQL ETL pipeline actually does before migrating it
- Prove a pipeline migration did not change its meaning using the evidence-backed Transformation IR
- Let a Claude or Cursor agent query a real repo through MCP with every answer citing a source fragment
- Cut token spend on repo questions by querying the compiled ledger instead of grepping raw files
- Generate component documentation for schema, write/read paths and migration frameworks
- Answer multi-step diagnostic questions such as why a metric spiked by chaining live MCP calls
- Point a local Ollama model at a ClickHouse server for NL-to-SQL without sending data to a hosted LLM
Setup time & first value
How long it actually takes to get something useful out of EKOS — broken out by persona, not the marketing-page minute.
Expect an afternoon, not a button: installing the CLI and getting `ekos build` through `ekos commit` working on your repo is the first milestone, and a 2,022-file cold ingest is timed at 34 seconds, so the wait is not the bottleneck — the recovery coverage is. Wiring `ekos mcp serve --workspace .` into Claude or Cursor is quick once the ledger exists. Teams that need to write custom analyzers for
Switching to or from EKOS
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From manual Pentaho job reading: run `ekos recover` over the job XML and read the recovered Source/Filter/Join/Aggregate/Sink nodes instead of opening each job file.
- →From ad-hoc grep over the repo: compile the ledger once, then query it over MCP so the same question costs 67-93% fewer tokens.
- →From a plain RAG index over repo files: rebuild the corpus through `ekos build` so answers cite content-addressable fragments rather than retrieved chunks.
- ↗To plain grep or ripgrep: nothing to export — the ledger is read-only and your source repo is untouched, though you lose the citations.
- ↗To a hosted code-search assistant: reindex against that vendor; EKOS compiles rather than indexes, so the Transformation IR does not port across.
- ↗To a custom in-house parser: the append-only evidence graph and CLI outputs are yours to keep, but you would be rewriting the compiler.
Integrations
Resources & guides
Tutorials & Learning
YouTube returned 6 videos for “EKOS”, and we withheld 6: 6 could not be judged, because “EKOS” is a single word that other videos use for other things. We are showing none, because we could not prove any of them are about EKOS.
Official links
What are the best EKOS alternatives?
The closest alternatives to EKOS we list are Roam Code, Chrome DevTools MCP and Pylar. Ranked by direct product-type match, not category overlap.
| Alternative | Pricing | Starts at | Last verified |
|---|---|---|---|
| Roam Code | Freemium | Pricing not published | — |
| Chrome DevTools MCP | Free | Free | — |
| Pylar | Freemium | Pricing not published | — |
Alternatives to EKOS
View all
Roam Code
Roam Code gives coding agents a queryable local code index plus free static checks that run without model calls.
Chrome DevTools MCP
Chrome DevTools MCP gives your AI coding agent a live Chrome for debugging, performance traces, and reliable browser automation.
Tools that pair well with EKOS
Common stack mates teams adopt alongside EKOS, with the specific reason each pairing earns its keep.
Roam Code
Roam Code gives coding agents a queryable local code index plus free static checks that run without model calls.
Chrome DevTools MCP
Chrome DevTools MCP gives your AI coding agent a live Chrome for debugging, performance traces, and reliable browser automation.
Pylar
Pylar turns SQL views into governed MCP tools so AI agents query only the data you approve.
EKOS vs its closest rival
EKOS's closest head-to-head is with Bito — If you're a developer who wants to quickly expose your GitHub repos to AI assistants with minimal fuss, EKOS is the fast, lightweight bridge. But if you're an enterprise team wrestling with multi-repo complexity and need architectural planning, impact analysis, and epic scoping, Bito's knowledge graph approach is the heavyweight contender. Choose based on your scale: solo/startup with a few repos → EKOS; multi-repo engineering org → Bito. Verdict generated
Head-to-head comparisons
Ekos vs Bito
If you're a developer who wants to quickly expose your GitHub repos to AI assistants with minimal fuss, EKOS is the fast, lightweight bridge. But if you're an enterprise team wrestling with multi-repo complexity and need architectural planning, impact analysis, and epic scoping, Bito's knowledge graph approach is the heavyweight contender. Choose based on your scale: solo/startup with a few repos → EKOS; multi-repo engineering org → Bito.
GeneratedEkos vs Smithery
If you want to supercharge your existing GitHub repos with AI, EKOS is the direct route. If you're building agents that need a universe of pre-built tools without auth headaches, Smithery is unbeatable. Choose based on whether you're repo-centric or agent-centric.
GeneratedFrequently asked questions
For makers of EKOS
- Get a Verified Listing for EKOS ($39): a written decision within 1 working day.
- Free listing for another tool: Free reviews are decided in order, within 3 weeks of the review starting.
Used EKOS? Help shape our editorial sentiment research.