Tenets
Free, local MCP server that uses NLP ranking to feed your AI coding assistant the right files instead of raw filesystem access.
If your assistant keeps grabbing the wrong files, Tenets attacks that directly: local NLP ranking — BM25, TF-IDF, import graphs, Git signals — replaces raw filesystem access, and pack-quality work like token-budget-aware packing with model-specific token counting is what makes it more than a grep wrapper. The bundled complexity, debt, and velocity analysis (tenets examine . --complexity --hotspots, tenets momentum --team) means you get code-health reporting from the same CLI. It costs nothing and your code stays put. Pass if you need hosted collaboration or a vendor service desk — those are Sourcegraph's territory, not this project's.
Verified 48m ago · liveness 64/100 · cite: rightaichoice.com/tools/tenets
- Developers in Cursor, Claude Desktop, or Windsurf who want smarter code context
- Privacy-conscious teams that cannot send source code to a cloud service
- Solo developers who want better AI code quality without API keys or spend
- Teams wanting consistent coding principles enforced across AI interactions
- Non-developers who need a GUI rather than a CLI and IDE config
- Teams needing cloud collaboration or shared multi-repo dashboards
- Enterprises requiring a vendor SLA or commercial support contract
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 Tenets if you need a GUI, shared multi-repo dashboards, or a vendor support contract — it is a local CLI and MCP server that has to be configured per IDE, and its collaboration model is your Git workflow rather than a hosted workspace.
Semantic embedding ranking is not in the core install — you have to add the tenets[ml] extra, which pulls in sentence-transformers and transformer dependencies and costs disk space and first-run download time.
Tenets is free and MIT-licensed, which puts it at the opposite end of the market from commercial code-intelligence platforms such as Sourcegraph — there is no seat cost, no per-repo charge, and no usage meter, because the processing is local. The trade is that you supply the machine, the install, and the support via documentation and community issues rather than a vendor service desk.
In short
Tenets — Free, local MCP server that uses NLP ranking to feed your AI coding assistant the right files instead of raw filesystem access. Best for Developers in Cursor, Claude Desktop, or Windsurf who want smarter code context, Privacy-conscious teams that cannot send source code to a cloud service, Solo developers who want better AI code quality without API keys or spend. Free to use.
What's new in Tenets
Checked todayAcross the latest 2 updates: 2 changelog entries.
Documentation hub launched
Tenets published a documentation hub with Quick Start, Supported Languages, CLI Reference, Configuration, Architecture, and API Reference sections to speed up onboarding.
Updated MCP config for Claude Desktop on macOS
Setup docs were corrected so Claude Desktop users on macOS are pointed to the config file at ~/Library/Application Support/Claude/.
What people actually say about Tenets — is it worth it?
We scanned public community sources for Tenets on Jul 3, 2026 and could not establish that the discussion we found is about this tool rather than something else sharing its name. Our own analysis of that scan says the posts were off-subject. Rather than publish a sentiment score built on the wrong subject, we publish nothing here and re-run the scan.
Viability Score
How well maintained and how widely used is Tenets? 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: October 2026
How we score →Key Features
- Multi-factor NLP code ranking (BM25, TF-IDF, keyword extraction, import graphs, Git signals)
- Native MCP server for Cursor, Claude Desktop, and Windsurf
- Local processing with no cloud API calls and no API keys for core ranking
- Open-source CLI plus Python library via pip install tenets[mcp]
- Optional local semantic embeddings and transformers via the tenets[ml] extra
- Token-budget-aware packing with model-specific token counting for GPT-4, Claude, and Llama
- Rules-based and ML summarizers built in; LLM summaries only with provider API keys enabled
- Persistent sessions with context branching and local SQLite storage
- Guiding principles (tenets) injection with priority and category to prevent LLM drift
- Cyclomatic, cognitive, and Halstead complexity metrics
- Maintainability index and code health scoring
- Technical debt tracking and test coverage analysis
- Anti-pattern and best-practice pattern detection
- Team velocity, contributor ownership, and hotspot analysis from Git history
- Interactive D3.js dependency graph visualization exported to HTML
About Tenets
Tenets is a free, open-source MCP server, Python library, and CLI that fixes the context problem in AI-assisted coding. Rather than giving your assistant raw filesystem access, it analyzes your repository with multi-factor NLP — BM25, TF-IDF, keyword extraction, import graphs, and Git signals — then ranks and packs the files that actually matter for a task you describe in plain English. You install it with pip install tenets[mcp] and add a single tenets-mcp line to your IDE's MCP config; Cursor, Claude Desktop, and Windsurf can then call its tools natively. Setup is documented for each of the three hosts, including the corrected macOS Claude Desktop config path at ~/Library/Application Support/Claude/. Everything runs on your machine: no cloud API calls, no code leaving your workstation, and no API keys required for core ranking. Alongside context building, Tenets ships a code analysis layer — cyclomatic, cognitive, and Halstead complexity metrics, maintainability and code health scoring, technical debt tracking, test coverage analysis, and anti-pattern detection. Development intelligence adds team and individual velocity (tenets momentum --team), hotspot analysis, contributor ownership, and interactive D3.js dependency graphs exported to HTML. Sessions persist context across prompts in local SQLite with branching, and 'tenets' — guiding principles you define with priority and category — get injected via tenets instill to keep AI behavior consistent and curb drift in long conversations. An optional tenets[ml] extra adds local sentence-transformer embeddings; OpenAI, Anthropic, and Cohere integrations stay off unless you enable them. AST parsing covers 15+ languages, and packing is token-budget aware with model-specific token counting for GPT-4, Claude, and Llama. It suits solo developers and privacy-conscious teams who want better AI code context without a cloud dependency. There is no hosted tier and no shared dashboard — for multi-repo policy or enterprise-grade code intelligence, a different class of commercial platform (Sourcegraph, for example) is the right shape of tool.
Behind the Verdict
Tenets makes a narrow, well-argued bet: the bottleneck in AI coding is not model quality but context selection, and you can solve a lot of it locally with classical IR instead of embeddings or a cloud index. The ranking pipeline is transparent — scan with .gitignore respected, run language analyzers and AST structure, then score with BM25, TF-IDF, keyword extraction, path relevance, import graph proximity, and Git churn signals. That combination is meaningfully better than the 'assistant reads the whole repo' default, and it costs no tokens to ship files that were never relevant. What distinguishes Tenets from a prompt wrapper is that the interesting work is not the LLM call — it is the analysis. Complexity metrics (cyclomatic, cognitive, Halstead), maintainability index and code health scoring, tech debt tracking, test coverage gaps, and anti-pattern detection all come from static analysis you can run without a model. The momentum and ownership views read Git history, so they work on any repo with real commits but degrade on shallow clones and squashed merges — worth knowing before you build a process on the numbers. The D3.js dependency graph (tenets viz deps --format html --output interactive.html) is the kind of artifact that is hard to get casually and genuinely useful before a rewrite. The honest weaknesses: it is a CLI plus IDE config, so it is not for anyone who wants a GUI. There is no shared session dashboard and no centralized multi-repo policy — collaboration happens through Git, not through the tool. Semantic ranking requires the separate tenets[ml] install. LLM-based summarization needs you to explicitly enable a provider API key; the default summarizers are rules-based and ML modes. And support is documentation plus community issues, not a service desk. The documentation hub published on 2025-09-01 (Quick Start, Supported Languages, CLI Reference, Configuration, Architecture, API Reference) closes a real onboarding gap for a tool whose first-run experience used to hinge on getting one JSON block right. Where it fits: developers in Cursor, Claude Desktop, or Windsurf on repos too large to paste, teams that cannot send source to a third party, and engineers who want complexity, debt, and velocity numbers out of the same binary. Where it does not: non-developers, orgs that need SSO-style central policy across many repos, and anyone who wants a managed service rather than a local install.
Researching Tenets? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Real-world workflow fit
Concrete scenarios for the personas Tenets actually fits — and what changes day-one when you adopt it.
Runs pip install tenets[mcp], adds the one-line tenets MCP entry to ~/.cursor/mcp.json, restarts Cursor, then calls tenets distill 'add mistral api to summarizer' instead of letting the assistant glob the repo.
Outcome: The assistant receives a ranked, token-budgeted set of files with rules-based summaries, so it edits the right module on the first pass rather than after two wrong-file attempts.
Runs tenets examine . --complexity --hotspots --ownership and tenets viz deps --format html --output interactive.html, then reviews the complexity metrics and D3.js coupling graph with the team.
Outcome: Refactor scope is chosen from cyclomatic/cognitive complexity, churn hotspots, and ownership data rather than from anecdote.
Creates a session (tenets session create payment-integration), adds tenets like 'Always validate user input' at critical/security and 'Use type hints in Python' at high/style, then runs tenets instill --session payment-integration before further prompts.
Outcome: Guiding principles are injected into the assistant's context from local SQLite, so behavior stays consistent across a long multi-prompt feature build.
Use Cases
- Rank and pack the exact files an AI assistant needs for a request like 'add mistral api to summarizer' (tenets distill).
- Define tenets such as 'Always validate user input' at critical priority and instill them so AI behavior stays consistent across long sessions.
- Run complexity, hotspot, and ownership analysis across a repo to plan a refactor (tenets examine . --complexity --hotspots --ownership).
- Generate an interactive D3.js dependency graph to understand module coupling before a rewrite.
- Track team velocity with tenets momentum --team --since 'last month' --detailed.
- Find untested code paths and technical debt to prioritize reliability work.
- Keep a persistent session going across many CLI invocations without restarting context.
- Add semantic similarity ranking on top of lexical scoring by installing the tenets[ml] extra.
Models Under the Hood
as of 2026-09-23
Limitations
- Tenets is entirely local, so there is no hosted tier, no shared session dashboard, and no centralized policy management across repositories — collaboration happens through your Git workflow, not the tool.
- Standalone use requires comfort with a CLI, and it is not aimed at non-developers.
- Semantic ranking needs a separate install (pip install tenets[ml]) beyond the core package.
- LLM-based summarization only kicks in if you explicitly enable a provider API key; the default summarizers are rules-based and ML modes.
- Project analysis and velocity features read Git history, so shallow clones or squashed commits limit ownership and hotspot accuracy.
- Support comes from documentation and community issues rather than a vendor service desk.
as of 2026-10-09
Verification history
We have re-verified Tenets 8 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.
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-checked, vendor evidence unchanged
- — re-checked, vendor evidence unchanged
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-checked, vendor evidence unchanged
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
Showing the 6 most recent of 8 verification passes.
Free to cite with attribution — this page re-verifies continuously.
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Plans compared
For each published Tenets tier: who it actually fits, and what it adds vs. the previous tier. Cross-reference the cost calculator above for projected annual outlay.
Free (open-source)
$0/mo
Ideal for
Solo developers and privacy-conscious teams running Cursor, Claude Desktop, or Windsurf who want better code context without a cloud dependency or API spend.
What this tier adds
Starting tier — the entire project is MIT-licensed and free, with local NLP ranking, analysis commands, and sessions included; semantic ranking requires the separate tenets[ml] extra.
Where the pricing makes sense
The company stage and team size where Tenets's pricing actually pencils out — and where peers do it cheaper.
Tenets is free and MIT-licensed, which puts it at the opposite end of the market from commercial code-intelligence platforms such as Sourcegraph — there is no seat cost, no per-repo charge, and no usage meter, because the processing is local. The trade is that you supply the machine, the install, and the support via documentation and community issues rather than a vendor service desk.
Setup time & first value
How long it actually takes to get something useful out of Tenets — broken out by persona, not the marketing-page minute.
Solo developer in Cursor: minutes — one pip install, one JSON block in ~/.cursor/mcp.json, restart the IDE. Claude Desktop adds a config-file location step (on macOS, ~/Library/Application Support/Claude/). Windsurf is Settings → Extensions → MCP. Getting real value beyond a first distill call takes longer: you need a repo worth ranking, and semantic ranking adds a second install (tenets[ml]).
Switching to or from Tenets
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From pasting files into chat manually: switch to tenets distill or tenets rank so the assistant pulls ranked files itself, and keep the CLI available for the same query outside the IDE.
- →From raw filesystem MCP access: keep the filesystem server for writes but route context selection through Tenets' ranking tools so the assistant stops globbing the whole repo.
- →From no session continuity: create a session (tenets session create <name>) and add tenets before your next multi-prompt feature so earlier context is not rebuilt from scratch.
- ↗To a hosted code-intelligence platform such as Sourcegraph: export nothing — re-point your IDE at the hosted service, and accept that per-repo policy and shared dashboards now come with a vendor relationship.
- ↗To raw filesystem MCP access: remove the tenets entry from your IDE's MCP config JSON and restart; your repo is untouched because Tenets only stored sessions in local SQLite.
Integrations
Resources & Guides
Tutorials & Learning
YouTube returned 6 videos for “Tenets”, and we withheld 6: 6 could not be judged, because “Tenets” 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 Tenets.
Official links
Tools that pair well with Tenets
Common stack mates teams adopt alongside Tenets, with the specific reason each pairing earns its keep.
Chrome DevTools MCP
Open-source MCP server that gives coding agents live Chrome DevTools access for debugging, automation, and performance traces.
Codeium
Devin Desktop is a free AI coding assistant and agent IDE with unlimited Tab completions and an Agent Command Center for fleets of local
Context7
Context7 feeds your AI coding assistant up-to-date, version-pinned library docs over MCP so it stops inventing APIs
Featured Head-to-Head Comparisons
Tenets vs Spider Cloud
Choose Tenets if you need private, local code context for AI coding assistants and are comfortable with CLI/API setup. Choose Spider Cloud if your AI agent requires real-time web data at scale, with a focus on scraping performance and cloud connectors.
Tenets vs Voyage Ai
Voyage AI is the clear choice for enterprises building domain-specific RAG pipelines that require high accuracy on finance, legal, or code data, with long-context embeddings and HIPAA/SOC 2 compliance. Tenets is ideal for privacy-conscious developers using AI coding assistants like Cursor or Claude Desktop, offering a free, open-source tool that improves context selection locally without cloud dependencies. Choose Voyage for retrieval scale and compliance; choose Tenets for code-level AI assistance and full data sovereignty.
Tenets vs Temporal Ai
If you need to build reliable, fault-tolerant AI agents that survive crashes and coordinate multi-step processes, Temporal AI is the clear choice with its durable execution engine and enterprise integrations. For developers who want to supercharge their local AI coding assistants with intelligent, privacy-preserving context selection, Tenets is a zero-cost, lightweight addition. They address different layers of the AI stack—choose based on whether your bottleneck is execution reliability or context quality.
Appgyver vs Tenets
If you're a developer using AI coding assistants like Cursor or Claude Desktop and want better context without sending code to the cloud, Tenets is the clear choice—it's free, local, and open-source. For SAP customers needing to build extensions or automate workflows within the SAP ecosystem, AppGyver (SAP Build) is the only option that offers clean-core compliance and deep integration with S/4HANA. They serve completely different purposes; choose based on your ecosystem.
Alternatives to Tenets
View all
Chrome DevTools MCP
Open-source MCP server that gives coding agents live Chrome DevTools access for debugging, automation, and performance traces.
Frequently Asked Questions
Best-of guides
Used Tenets? Help shape our editorial sentiment research.