# AXLR > Open-source Go execution engine for agentic work from Underpass AI. A console with work modes and approval cards, an mTLS HTTP, gRPC and MCP service, a JSON worker and a Go library, with KMP memory and MADE procedures over MCP. - Canonical: https://www.anchorterminal.com/tools/axlr - Pricing: https://www.anchorterminal.com/tools/axlr#pricing - Reviews: https://www.anchorterminal.com/tools/axlr#reviews - Markdown: https://www.anchorterminal.com/tools/axlr.md (~10,000 tokens) - Slim: https://www.anchorterminal.com/tools/axlr.min.md (~1,780 tokens, same facts, less prose, for token-sensitive contexts) - JSON: https://www.anchorterminal.com/tools/axlr.json (this page as data, same URL with Accept: application/json) - Site index for agents: https://www.anchorterminal.com/llms.txt (the key docs and every listing in one file, at most 16 MB, about 4 million tokens: https://www.anchorterminal.com/llms-full.txt) - API: https://www.anchorterminal.com/api/v1/index.json - MCP: https://www.anchorterminal.com/mcp (the whole catalogue as tools, no key) - DNS: dig +short TXT axlr.grade.anchorterminal.com (the grade as a DNS TXT record, for a sandbox with no HTTP where names still resolve; also top., best.. and .vs.. under the zone, and the same under grade.letme.dev) - Licence: CC BY 4.0, credit Anchor Terminal and link this page (https://www.anchorterminal.com/terms/#data) - Updated: 2026-11-01 ## Overview **Grade BB · 70.6/100 · rank #163 of 1253 · #3 in Agent harnesses · agent-ready · confidence medium** ## Assessment A pre-1.0 Go harness with an unusually complete service contract, mutual TLS with principal roles, required idempotency keys, a policy, budget and sandbox per execution, and a result object a program can read. The console's defaults after setup are autonomy on and no sandbox, the project has one maintainer and no outside users yet, and releases carry no notes. ## Facts | Field | Value | | --- | --- | | Vendor | Underpass AI (https://axlr.underpassai.com) | | Kind | Agent harness | | Category | Agent harnesses (https://www.anchorterminal.com/categories/agent-harnesses) | | Auth | OAuth or key · The console and the one-command `axlr-tui --exec` need a model route, an OpenRouter key (`OPENROUTER_API_KEY` or `credentials.json`, mode 0600), a local OpenAI-compatible server, or an Anthropic Messages or OpenAI Responses key in a `local_models` entry; the JSON worker and the library tests need no model account. The service, `axlr-serve`, requires TLS 1.3 and a client certificate signed by its CA whose SHA-256 fingerprint is mapped in a private principals file to a principal and the roles `session_client`, `approver`, `tool_operator` and `admin`; an unmapped certificate is forbidden, a session belongs to the principal that created it, and authentication never approves a tool by itself. MCP at `/mcp` is stateless and authenticated per request by the same certificate; gRPC uses the TLS peer certificate. | | Pricing | Free (Free · OSS) · Free and open source under Apache-2.0, with nothing to buy from Underpass AI. You pay your model provider (OpenRouter reports a cost per request, which the console's `/usage` and `max_session_usd` track) or run a local server at no charge; TypeSafe's Jev judge is optional and needs its own key. No hosted AXLR exists (checked 2026-11-01). | | x402 | No · No x402, MPP or L402 in the repository, the OpenAPI file or the two sites (checked 2026-11-01). Nothing is sold. | | Licence | Apache-2.0, with a NOTICE naming Tirso García Ibáñez as the copyright holder | | Packages | go: `github.com/underpass-ai/AXLR`; oci: `ghcr.io/underpass-ai/axlr`; oci: `ghcr.io/underpass-ai/charts/axlr` | | Source | https://github.com/underpass-ai/AXLR | | Docs | https://axlr.underpassai.com/docs/ | | llms.txt | not found | | Last release | 2026-10-25 | | GitHub stars | 0 (as of 2026-11-01) | | Interfaces | axlr-tui console (interactive, `--exec` headless, `setup`, `skills`, `logs`), axlr-serve (HTTP /v1, gRPC, Streamable HTTP MCP at /mcp, mTLS, Helm chart), axlr JSON worker (one request on stdin), Go library | | Built-in tools | local_read, local_write, local_edit, local_exec, local_search, local_list; host tools for sessions, memory, skills, catalog and forged tools; KMP and MADE over MCP | | Approvals | Cards per call with always-allow and `/autonomy`; the setup assistant writes autonomy on with a destructive list that always asks. `--approve none\|local\|all` and a policy level per execution. The service approves nothing but reads | | Sandbox | None by default; the console and worker are stated not to be sandboxes. `exec_sandbox` (bubblewrap, Linux) in settings, or a `process` or `container` profile per execution with hidden homes and a private /tmp | | Network | Commands get the host's network unless the sandbox egress is `none` or an `allowlist` enforced through AXLR's own HTTP proxy; `routing.allow_remote_providers: false` keeps model requests on loopback | | MCP client | stdio and Streamable HTTP, manual or automatic approval per server, Codex-compatible plugin packages; no OAuth sign-in or custom headers | | Models | OpenRouter, local OpenAI-compatible servers, Anthropic Messages and OpenAI Responses APIs directly; a routing per execution ranks candidates by cost, latency or quality | | Headless | `axlr-tui --exec --json` with a result object and exit codes 0, 3 and 1; the service's CreateSession, StartTurn, StreamEvents and approvals over HTTP, gRPC or MCP | | Telemetry | None found. Model requests go to the chosen provider; Jev to TypeSafe only when on; engine downloads from api.github.com; catalog index from registry.underpassai.com. Payload capture off by default, traces kept 30 days | | Releases in 90 days | 21 (v0.1.0 on 4 October to v0.15.0 on 25 October 2026), six archives each with .sha256, a Helm chart and an image digest; empty release notes | | Capabilities | agent.harness, agent.mcp-client, agent.multi-agent | | Tags | open-source, local, self-hosted, free, no-card, go, cli, mcp-client, openapi, grpc, docker, helm, sandbox, approvals, pre-1.0, new | | JSON | https://www.anchorterminal.com/api/v1/tools/axlr.json | ## Score breakdown (methodology v0.7, October 2026 research run) Assessed 2026-11-01 from public evidence against the published checklist (https://www.anchorterminal.com/benchmark/#checklist). Confidence: medium. Performance and Task success pending (no score, not in the total); the total is Σ(score × weight) ÷ 80 over the 7 assessed categories. "This run" is each category's share of the 100 points. | Category | Weight | This run | Score (0–100) | Points | | --- | --- | --- | --- | --- | | Reliability | 16% | 20 | 70 | 14.0 | | Performance | 10% | pending | pending | n/a | | Schema & documentation | 13% | 16.2 | 80 | 13.0 | | Agent ergonomics | 13% | 16.2 | 83 | 13.5 | | Security & auth | 14% | 17.5 | 62 | 10.8 | | Payments & pricing | 10% | 12.5 | 60 | 7.5 | | Task success | 10% | pending | pending | n/a | | Maintenance & community | 7% | 8.8 | 79 | 6.9 | | Transparency & trust (editorial 68, provenance 42) | 7% | 8.8 | 55 | 4.8 | | Negative events | up to −15 | up to −15 | none recorded | 0 | | **Total** | | | | **70.6 → BB** | ### Why each score - Reliability 70 (checked 2026-11-01): Read as a local package, as claude-code, goose, opencode and openhands were. Release archives for Linux, macOS and Windows on amd64 and arm64 with a `.sha256` beside each, a GHCR image, an OCI Helm chart and source builds with Go 1.26, all stated in the docs; no Homebrew, npm or apt route (20). CI on main runs vet, race tests and a coverage gate above 80 per cent for both modules, with the sandbox tests required under bubblewrap, and the last ten runs on main to 1 November 2026 all passed; the release workflow tests on six native runners (25). 10 open issues, all opened by the founder on 19 and 21 October 2026 as trackers for service work (work modes, plugins, the sandbox setting, cost limits), none from an outside user and none a crash report, 0 open pull requests and about 100 merged; with 0 stars there is nobody else to report, so the handling of outside reports is untested (20). Tags are semver 0.x and the docs date each behaviour change inline ("until 14 October 2026…"), but the 21 releases have empty notes, there is no CHANGELOG.md, and the product contract records one retirement (the seven 1.0 skill definitions on 5 October 2026) without a notice period (5). Pre-1.0, which the README calls an evolving project (0). 20+25+20+5+0 = 70. - Performance: Pending. Latency is measured per call by our probes, which haven't run yet, so this run doesn't score it. Its weight is shared across the assessed categories until the first probe window closes. - Schema & documentation 80 (checked 2026-11-01): An OpenAPI 3.1 file for the service (15 paths, 23 schemas, `mutualTLS` security scheme, `additionalProperties: false` on request bodies), a checked proto for the gRPC service, and the 19 MCP tools built from the same shared schemas (25). No llms.txt on either site (404), and the docs site renders HTML; the Markdown sources are in the repository (5). The product contract tables what each of the four interfaces can and cannot do, the work-mode table states each mode's purpose and tool policy, the transport parity ledger names every verb's arguments and behaviour, and `axlr_tools` tells the model to look for a tool before forging one (15). Typed inputs throughout, `Idempotency-Key` 16 to 128 safe ASCII characters, `expected_revision` a positive integer, policy levels, sandbox profiles and egress values enumerated, budgets as four integers, and the service rejects unknown fields, invalid types and bodies over 4 MiB (14). Examples in every guide, an error shape of `code`, `message` and `request_id`, named codes (`409 idempotency_conflict`, `400 invalid_cursor`, `422 invalid_policy`, `invalid_budget`, `sandbox_unavailable`, `not_ready`, `invalid_id`), a gRPC status mapping, and the worker's five statuses and three exit codes; no single error reference page (14). `/v1` routes and an OpenAPI version of 0.1.0 against a 0.15.0 release, and no changelog beyond the dated notes inside the docs (7). 25+5+15+14+14+7 = 80. - Agent ergonomics 83 (checked 2026-11-01): Harness reading of the framework line, scored on what a program driving it has to supply. Six local tools plus the host's bookkeeping tools; work modes and an execution policy hide the tools a run may not use, `axlr_catalog` searches the catalog instead of loading it, forged tools keep the request prefix for the prompt cache, and a context cap of 75,000 prompt tokens opens a successor session (20). `local_read`, `local_search` and `local_list` page with `offset`, `next_offset` and `max_bytes`; `StreamEvents` takes `after`, `max_events` (1 to 256) and `wait_ms`; a budget per execution in seconds, requests, tool calls and dollars; 32 tool calls a turn and `max_session_usd` (20). The `--exec` run ends in one `result` line with a status, the pending call, every call's outcome and the usage, and exits 0, 3 or 1; service errors carry codes an agent can act on; a budget stop says which limit (`budget: requests 20 reached`) and leaves the session resumable (16). `Idempotency-Key` on every turn, approval and direct call with cross-transport replay, `expected_revision` on decisions, `expected_sha256` on worker writes, and an interrupted effectful call left `uncertain` for an operator rather than retried (19). One command runs a task, but the first launch needs an OpenRouter key or a local server, the service needs a CA, a principals file and remote KMP and MADE before a first session, and the only SDK is the Go library (8). 20+20+16+19+8 = 83. - Security & auth 62 (checked 2026-11-01): Harness reading of the framework checklist, as for the other four harnesses. 30 for what leaves the machine by default, 20 for approvals and sandboxing, 15 for prompt-injection posture, 15 for audit and 20 for the security programme. No product telemetry in the console source (the outbound hosts are openrouter.ai, api.anthropic.com, api.openai.com, a configured local server, api.github.com and github.com for engine downloads, and registry.underpassai.com for the signed catalog index) or in the docs; payload capture is off by default since 9 October 2026, traces omit prompts and credentials, model keys are never given to commands or plugins, and Jev sends requests and answers to TypeSafe only when on, which the setup assistant does when `TYPESAFE_API_KEY` is set, with a notice on stderr (27). Approval cards, a destructive list that always asks, four work modes, a policy level per execution, a bubblewrap or container sandbox with an enforced egress allow list, and a service that approves nothing but reads. Against that, the setup assistant writes `autonomous: true`, `exec_sandbox` is off by default, forged tools run with the person's rights, and the README says the console and worker are not sandboxes (13). Forged tools are refused when a file changes since the forge, a policy forbids calls outside the model, and the docs say MCP tools keep their own side effects, but no prompt-injection guidance was found and MCP results reach the model unmarked (6). The service keeps an event journal, audit records, a receipt with the SHA-256 of every call's arguments and outcome, and correlation ids on every event; the console keeps session records, a redacted app log and `/session` (14). No SECURITY.md, no security.txt on either host, no advisories, no bug bounty; release checksums and a minisign-signed catalog are integrity, not a programme (2). 27+13+6+14+2 = 62. - Payments & pricing 60 (checked 2026-11-01): No payment protocol (0). Free and Apache-2.0 with nothing to buy, so 20, 20 and 20 on the last three lines by the self-hosted rule, since the docs say what you pay your model provider, the free route needs no card, and it runs local models with no signup at all, the JSON worker without any model. 0+20+20+20 = 60. - Task success: Pending. Task success needs the category task suites run through each tool, which haven't run yet, so this run doesn't score it. Its weight is shared across the assessed categories until then. A data provider's data-quality score is published on its listing now and becomes half of this category when it's scored. - Maintenance & community 79 (checked 2026-11-01): v0.15.0 on 2026-10-25, with commits on main to 2026-11-01 (30). 21 releases between 4 and 25 October 2026 (20). One maintainer (368 of 399 commits, the rest by two agent accounts), 10 open issues that are his own trackers with no comments, about 100 merged pull requests and no outside report to measure a reply against (15). Not in the MCP registry; the Go module and a proto to generate clients from are the only client code, beside the GHCR image and Helm chart (5). Go 1.26 toolchain pinned, release actions pinned by commit SHA, a go-licenses check that the third-party bundle matches the dependency graph, and CI passing on main (9). 30+20+15+5+9 = 79. - Transparency & trust 55 (checked 2026-11-01): Apache-2.0 with a NOTICE (30). No privacy policy or terms on underpassai.com or axlr.underpassai.com; the docs' 'Where data goes' and 'Diagnostics and privacy' sections say that model requests go to OpenRouter, the local server or the native API, that Jev sends text to TypeSafe only when on, that sessions live under `$XDG_STATE_HOME/axlr`, that payload capture is off by default and redacted when on, and that traces are deleted after 30 days; nothing covers the engine downloads' or the catalog registry's handling of a client's requests (15). No deprecation policy; pre-1.0 changes are dated inline in the docs and the product contract records the 5 October 2026 retirement of the first skill catalogue after the fact (6). No telemetry to disclose, and the docs say in two places what leaves the machine and how to stop payload capture, though no single page says there is no telemetry (17). 30+15+6+17 = 68. Fix list, everything this grade says the listing lacks, the biggest gain first (19 items). What it covers is at https://www.anchorterminal.com/fixes/axlr.md (JSON https://www.anchorterminal.com/fixes/axlr.json). The whole list for a coding agent comes from `POST https://www.anchorterminal.com/api/v1/fix-list` with `{"listing": "axlr", "plan": "read"}`, free while the beta runs ### What we couldn't check - Unchecked: nothing was built or run, so the setup assistant, the sandbox, the service and the exit codes are described from the source and docs, not observed - Unchecked: registry.underpassai.com answers every path, robots.txt included, with 403 'forbidden: client is not AXLR', so the signed index, its minisign signature and the 159-tool count are the vendor's statements - Unchecked: the console's engine updater names `kmp-plugin--.tar.gz` and `made-plugin-…` assets in the engines' latest releases; the KMP v0.25.0 asset list we read showed binaries and no such archive in its first twelve entries, so whether the console's KMP install succeeds today was not established - The OpenAPI file is versioned 0.1.0 while the release is 0.15.0, so it may lag the code - The underpassai.com components section says AXLR is at 0.14.0 while axlr.underpassai.com and GitHub say v0.15.0 - No outside user has filed an issue, so the maintainer's responsiveness to reports is unknown - Whether the native Anthropic and OpenAI adapters work against the live services; the docs say their shapes were checked against provider documentation and not the services - The legal entity is absent; the site says incorporation as a Spanish limited liability company is being prepared, with no date ### Sources - repository README (trusted workspace, not a sandbox, project status): (seen 2026-11-01) - product contract (interface table, guarantees and limits, audited 4 October 2026): (seen 2026-11-01) - service API (mTLS, principals, roles, idempotency, errors, policy, budget, sandbox): (seen 2026-11-01) - transport parity contract (19 verbs over HTTP, gRPC and MCP): (seen 2026-11-01) - OpenAPI 3.1 file (15 paths, mutualTLS scheme, error schema): (seen 2026-11-01) - runtime contract (policy, budget, sandbox and routing per execution, result, receipt): (seen 2026-11-01) - getting started (setup assistant, where data goes): (seen 2026-11-01) - console guide (--exec and exit codes, work modes, exec sandbox, forged and catalog tools, Jev, diagnostics and privacy): (seen 2026-11-01) - configuration reference (approvals and destructive list, exec_sandbox, catalogs, native APIs, jev, credentials): (seen 2026-11-01) - setup defaults in source (CompleteDefaults writes autonomy on): (seen 2026-11-01) - engine updater source (latest release from api.github.com, .sha256 check): (seen 2026-11-01) - engine lock for the service image (pinned versions and SHA-256): (seen 2026-11-01) - worker contract (tools, limits, statuses, exit codes): (seen 2026-11-01) - release process: (seen 2026-11-01) - release workflow (six native runners, checksum verification, publish without replacing bytes): (seen 2026-11-01) - CI workflow (race tests, coverage gate, sandbox tests): (seen 2026-11-01) - LICENSE and NOTICE: (seen 2026-11-01) - releases (21, assets, empty notes): (seen 2026-11-01) - repository metadata (created 2026-09-29, 0 stars, 10 open issues): (seen 2026-11-01) - open issues: (seen 2026-11-01) - CI runs on main: (seen 2026-11-01) - organisation (created 2025-08-09, Spain, 16 repositories): (seen 2026-11-01) - a release checksum asset: (seen 2026-11-01) - AXLR site home (v0.15.0, forged tools, 159 catalog tools): (seen 2026-11-01) - install page (download and verify, setup, autonomy after setup, sandbox off): (seen 2026-11-01) - Underpass AI home (components, founder, incorporation note, contact): (seen 2026-11-01) - security.txt (404): (seen 2026-11-01) - RDAP for underpassai.com: (seen 2026-11-01) ## Score history Every move of a score since the listing was graded, newest first, with its cause. A score can't move without an entry here. JSON: https://www.anchorterminal.com/api/v1/tools/axlr/history.json - 2026-11-01: graded 70.6 BB (methodology 0.7) ## Who's behind it (provenance 42/100, checked 2026-11-01) | Check | Finding | Points | | --- | --- | --- | | Legal entity named | not found | 0/20 | | Domain age | underpassai.com, registered 2025-08-10 (1 year) | 3/15 | | Endpoint on the vendor's domain | no hosted endpoint | n/a | | Terms of service | nothing hosted, so the Apache-2.0, with a NOTICE naming Tirso García Ibáñez as the copyright holder licence stands in | 10/10 | | Privacy policy | nothing hosted, not scored | n/a | | Status page | not found | 0/10 | | Changelog | published | 10/10 | | security.txt | not found | 0/10 | | People named | 1 named with their role, 1 with a public profile | 10/10 | | Support route | a support address, help pages, a community | 5/5 | No legal entity exists yet. underpassai.com's footer says Underpass AI is a startup project created by Tirso García Ibáñez and preparing incorporation as a Spanish limited liability company, that the technical assets are founder-owned and intended to be assigned to the company after incorporation, and that licence terms are in each repository's LICENSE and NOTICE. The NOTICE names Tirso García Ibáñez as copyright holder. Recorded as absent with that explanation. No terms of service or privacy policy on underpassai.com (sitemap lists the home page and one article) or axlr.underpassai.com. For local software the docs' 'Where data goes' and 'Diagnostics and privacy' sections stand in. RDAP for underpassai.com gives registration on 2025-08-10 with Route 53 nameservers. The GitHub organisation underpass-ai was created on 2025-08-09, location Spain, 16 public repositories. The AXLR repository was created on 2026-09-29; the first tag and release, v0.1.0, is dated 2026-10-04, which is firstReleased. https://underpassai.com/.well-known/security.txt and https://axlr.underpassai.com/.well-known/security.txt answer 404; the repository has no SECURITY.md. No status page (local software). The changelog is the GitHub releases list, whose entries carry no notes. robots.txt on underpassai.com and axlr.underpassai.com allow everything. registry.underpassai.com answers every path, robots.txt included, with 403 'forbidden: client is not AXLR', so nothing was read from it after that reply. The founder's LinkedIn URL is linked from underpassai.com and was not visited. ### Terms and privacy, as read A reading by a fixed set of rules, each answered with the vendor's own sentence. Not legal advice. **Terms of service**. Nothing is hosted by the vendor, so there are no terms of service to read. The Apache-2.0, with a NOTICE naming Tirso García Ibáñez as the copyright holder licence stands in and the check scores in full. **Privacy policy**. Nothing is hosted by the vendor, so there is no privacy policy to read and the check isn't scored. ## Humans behind the project (researched 2026-11-01 from public pages, not part of the grade) ### Key people - Tirso García Ibáñez, Founder, Underpass AI. LinkedIn: https://www.linkedin.com/in/tirsogarcia/. GitHub: https://github.com/tgarciai - Note: underpassai.com's Founder section says Underpass AI is created by Tirso García Ibáñez, a software architect, and its footer says the project is preparing incorporation as a Spanish limited liability company with founder-owned assets to be assigned to the company afterwards. The AXLR NOTICE names him as copyright holder, and the GitHub user tgarciai made 368 of the repository's 399 commits (the other 31 are by two agent accounts). The LinkedIn URL is the one the company site links and was not visited. Nobody else is named on either site or in the repository. Ownership is recorded as a project because no company exists yet, from the site's own footer. Investors were looked for on underpassai.com (home page, sitemap of two pages, articles section), the GitHub organisation profile and six web searches for the company, the founder, funding, a round and news coverage, which returned nothing about Underpass AI; no Y Combinator page, filing or press report was found. - Underpass AI: GitHub https://github.com/underpass-ai ### Getting support - Support email: contact@underpassai.com - Help: https://axlr.underpassai.com/docs/ - Help: https://github.com/underpass-ai/AXLR/blob/main/docs/troubleshooting.md - Community (GitHub issues): https://github.com/underpass-ai/AXLR/issues - Sources: https://underpassai.com/, https://underpassai.com/sitemap-0.xml, https://axlr.underpassai.com/, https://axlr.underpassai.com/docs/, https://github.com/underpass-ai/AXLR/blob/main/NOTICE, https://api.github.com/repos/underpass-ai/AXLR/contributors, https://api.github.com/orgs/underpass-ai ## Probe metrics Not measured yet. Our benchmark probes haven't run, so there's no availability, latency or error rate from a run and Performance is pending. Live uptime, where we poll the endpoint, is under Live and doesn't change the score. ## Strengths - mTLS HTTP, gRPC and MCP service with four principal roles, `Idempotency-Key` required on turns, approvals and direct calls, and `expected_revision` on decisions - A policy (observe, propose, act-with-approval, act-within-policy), a budget in seconds, requests, tool calls and dollars, and a bubblewrap or container sandbox with an enforced egress allow list, each per execution - `axlr-tui --exec --json` runs one task headless and ends with a result object; exit 0 done, 3 waiting for a decision, 1 error - OpenRouter, local OpenAI-compatible servers (llama.cpp, vLLM, Ollama, LM Studio), and the Anthropic Messages and OpenAI Responses APIs spoken directly - 21 releases in October 2026 from a workflow that tests on six native runners and refuses to republish different bytes, with a `.sha256` beside every archive ## Weaknesses - The setup assistant turns autonomy on and leaves the exec sandbox off, so after setup the model's commands run with the person's rights, bar the destructive list - Pre-1.0 with one maintainer, 0 stars and 10 open issues, all the founder's own trackers; releases have empty notes and there is no changelog - The service needs KMP and MADE as remote gRPC services, a client CA and a principals file before a first session, and has no work modes, plugins or forged-tool setting yet - No SECURITY.md, security.txt, advisories or disclosure route; the legal entity is not yet incorporated and no terms or privacy policy exist - No llms.txt, and the Go library is the only SDK; the catalog registry answers every non-AXLR client with 403 ## Before you call it (notes for agents) 1. Before the first prompt, set `approvals.autonomous` to false or run with `--approve none`, and turn `exec_sandbox.mode` to `auto` or `required` on Linux; after setup both default the other way 2. Drive a run with `axlr-tui --exec --json --policy --budget --sandbox ` and read only the last `result` line; exit 3 means a call waits for `--decide approve|deny` with `--session ` 3. On the service, send `Idempotency-Key` (16 to 128 safe ASCII characters) on every turn, approval and direct call, and pass the session's `expected_revision`; a reused key with a different body answers 409 4. A local model must return native tool calls (llama.cpp with `--jinja`, vLLM with `--enable-auto-tool-choice`); set `context_tokens` on the entry and `OLLAMA_CONTEXT_LENGTH` for Ollama 5. Leave `TYPESAFE_API_KEY` unset unless you want the setup assistant to switch Jev on, which sends requests and final answers to TypeSafe ## Connect Install: ```bash V=v0.15.0; OS=linux; ARCH=amd64; curl -fsSLO https://github.com/underpass-ai/AXLR/releases/download/$V/axlr-$V-$OS-$ARCH.tar.gz{,.sha256} && sha256sum -c axlr-$V-$OS-$ARCH.tar.gz.sha256 && tar xzf axlr-$V-$OS-$ARCH.tar.gz -C ~/.local/bin axlr axlr-tui axlr-serve # or build from source with Go 1.26 ``` Headless / CI: ```json { "command": "axlr-tui --root \"$PWD\" --exec \"$TASK\" --json --approve none --policy policy.json --budget budget.json --sandbox sandbox.json", "decide": "axlr-tui --session \u003cid\u003e --decide approve --json", "env": { "OPENROUTER_API_KEY": "\u003ckey\u003e" } } ``` ## Similar tools Ranked by shared capabilities, then score. Same-category tools with no shared capability key are listed last. | Tool | Grade | Score | Rank | Shared capabilities | x402 | Markdown | | --- | --- | --- | --- | --- | --- | --- | | goose | BB | 71.3 | 138 | agent.harness, agent.mcp-client, agent.multi-agent | no | https://www.anchorterminal.com/tools/goose.md | | OpenHands | BB | 70.3 | 169 | agent.harness, agent.mcp-client, agent.multi-agent | no | https://www.anchorterminal.com/tools/openhands.md | | OpenCode | B | 68.8 | 198 | agent.harness, agent.mcp-client, agent.multi-agent | no | https://www.anchorterminal.com/tools/opencode.md | | Qwen Code | B | 68.7 | 202 | agent.harness, agent.mcp-client, agent.multi-agent | no | https://www.anchorterminal.com/tools/qwen-code.md | | Kilo Code CLI | B | 68 | 243 | agent.harness, agent.mcp-client, agent.multi-agent | no | https://www.anchorterminal.com/tools/kilo-code-cli.md | | MiniMax Code | B | 67.7 | 258 | agent.harness, agent.mcp-client, agent.multi-agent | no | https://www.anchorterminal.com/tools/minimax-code.md | ## Panel reviews (0) Reviewed by the Anchor panel (https://www.anchorterminal.com/reviewers/index.md): . Desk reviews, written from public documentation, pricing, terms, source and status history on 1 October 2026. No calls made. For a desk review, the outcome says whether the reviewer's questions could be answered from public material: success, partial or failure. How reviews work: https://www.anchorterminal.com/reviews/how-it-works.md ## Notable - The setup assistant on first launch writes `settings.json` with autonomy on, so every tool the work mode permits runs without a card, except the destructive list (deletions outside the workspace, git push, reset --hard, clean, rebase, branch deletion and network writes such as `kubectl apply` or `docker push`), which always asks. `exec_sandbox` is off by default, so commands and forged tools run with the person's rights (source: ) - The README and the product contract say the console and the JSON worker are not sandboxes. A work mode constrains the local file tools and process approval but does not sandbox programs or classify MCP side effects; `exec` and MCP servers use the host account's permissions (source: ) - The same setup downloads the latest KMP and MADE releases from api.github.com on Linux and Apple-silicon macOS, checks each archive against the `.sha256` published beside it in the same release, and registers both as MCP servers with automatic approval; the service image pins both engines by version and SHA-256 in distribution/engines.lock.json (tui/adapters/engineupdate/updater.go) - The model can forge a tool in the session (`axlr_forge_tool` writes one to eight files under `.axlr/tools//`, approved as a write, run as a `local_exec` with the file hashes recorded, refused if a file changes) or install one of 159 pre-forged tools from a minisign-signed catalog at registry.underpassai.com (source: ) - 21 tagged releases between 4 and 25 October 2026 (v0.1.0 to v0.15.0), each with six archives for Linux, macOS and Windows on amd64 and arm64, a `.sha256` beside each, a Helm chart and an image digest; the release notes are empty and there is no CHANGELOG.md (source: ) - No SECURITY.md, no security.txt on either site, no published advisories. No product telemetry was found in the console source or docs; payload capture is off by default and Jev sends text to TypeSafe only when switched on, with a notice on stderr - #3 of 46 in Best agent harnesses and coding agents: https://www.anchorterminal.com/best/agent-harnesses/index.md - All 1000 harnesses comparisons: https://www.anchorterminal.com/compare/agent-harnesses/index.md ## Compare - [Agent Zero vs AXLR](https://www.anchorterminal.com/compare/agent-zero-vs-axlr.md): C 54.6 vs BB 70.6 - [Aider vs AXLR](https://www.anchorterminal.com/compare/aider-vs-axlr.md): D 47.1 vs BB 70.6 - [Amp vs AXLR](https://www.anchorterminal.com/compare/amp-vs-axlr.md): C 57.1 vs BB 70.6 - [Antigravity CLI vs AXLR](https://www.anchorterminal.com/compare/antigravity-cli-vs-axlr.md): C 57.6 vs BB 70.6 - [Auggie CLI vs AXLR](https://www.anchorterminal.com/compare/auggie-vs-axlr.md): D 51 vs BB 70.6 - [AXLR vs Claude Code](https://www.anchorterminal.com/compare/axlr-vs-claude-code.md): BB 70.6 vs C 59.9 - [AXLR vs Cline](https://www.anchorterminal.com/compare/axlr-vs-cline.md): BB 70.6 vs C 60.5 - [AXLR vs Crush](https://www.anchorterminal.com/compare/axlr-vs-crush.md): BB 70.6 vs C 56.7 - [AXLR vs Cursor CLI](https://www.anchorterminal.com/compare/axlr-vs-cursor-cli.md): BB 70.6 vs F 35.4 - [AXLR vs DeepSeek Harness (dsh)](https://www.anchorterminal.com/compare/axlr-vs-deepseek-harness.md): BB 70.6 vs D 49.5 - [AXLR vs dsh-superagent (DeepSeek)](https://www.anchorterminal.com/compare/axlr-vs-deepseek-superagent.md): BB 70.6 vs E 42.1 - [AXLR vs Devin](https://www.anchorterminal.com/compare/axlr-vs-devin.md): BB 70.6 vs C 55.8 - [AXLR vs Droid CLI](https://www.anchorterminal.com/compare/axlr-vs-droid-cli.md): BB 70.6 vs C 59.2 - [AXLR vs Pi](https://www.anchorterminal.com/compare/axlr-vs-earendil-pi.md): BB 70.6 vs B 62.6 - [AXLR vs Gemini CLI](https://www.anchorterminal.com/compare/axlr-vs-gemini-cli.md): BB 70.6 vs B 66.8 - [AXLR vs GitHub Copilot CLI](https://www.anchorterminal.com/compare/axlr-vs-github-copilot-cli.md): BB 70.6 vs C 58.4 - [AXLR vs goose](https://www.anchorterminal.com/compare/axlr-vs-goose.md): BB 70.6 vs BB 71.3 - [AXLR vs Harness²](https://www.anchorterminal.com/compare/axlr-vs-harness2.md): BB 70.6 vs E 42 - [AXLR vs HarnessRouter](https://www.anchorterminal.com/compare/axlr-vs-harnessrouter.md): BB 70.6 vs D 53.3 - [AXLR vs Jules](https://www.anchorterminal.com/compare/axlr-vs-jules.md): BB 70.6 vs E 44.7 - [AXLR vs Junie CLI](https://www.anchorterminal.com/compare/axlr-vs-junie.md): BB 70.6 vs D 53.1 - [AXLR vs Kilo Code CLI](https://www.anchorterminal.com/compare/axlr-vs-kilo-code-cli.md): BB 70.6 vs B 68 - [AXLR vs Kimi Code CLI](https://www.anchorterminal.com/compare/axlr-vs-kimi-code-cli.md): BB 70.6 vs B 62.1 - [AXLR vs Kiro CLI](https://www.anchorterminal.com/compare/axlr-vs-kiro-cli.md): BB 70.6 vs C 60.8 - [AXLR vs mini-swe-agent](https://www.anchorterminal.com/compare/axlr-vs-mini-swe-agent.md): BB 70.6 vs C 57.1 - [AXLR vs MiniMax Code](https://www.anchorterminal.com/compare/axlr-vs-minimax-code.md): BB 70.6 vs B 67.7 - [AXLR vs Mistral Vibe CLI](https://www.anchorterminal.com/compare/axlr-vs-mistral-vibe-cli.md): BB 70.6 vs B 63 - [AXLR vs Muse Code](https://www.anchorterminal.com/compare/axlr-vs-muse-code.md): BB 70.6 vs C 61.2 - [AXLR vs OpenAI Codex](https://www.anchorterminal.com/compare/axlr-vs-openai-codex.md): BB 70.6 vs BB 73.5 - [AXLR vs OpenCode](https://www.anchorterminal.com/compare/axlr-vs-opencode.md): BB 70.6 vs B 68.8 - [AXLR vs OpenHands](https://www.anchorterminal.com/compare/axlr-vs-openhands.md): BB 70.6 vs BB 70.3 - [AXLR vs OpenMausBot](https://www.anchorterminal.com/compare/axlr-vs-openmausbot.md): BB 70.6 vs B 65.4 - [AXLR vs Prime Agent](https://www.anchorterminal.com/compare/axlr-vs-prime-agent.md): BB 70.6 vs C 60.7 - [AXLR vs Qoder CLI](https://www.anchorterminal.com/compare/axlr-vs-qoder-cli.md): BB 70.6 vs C 59 - [AXLR vs Qwen Code](https://www.anchorterminal.com/compare/axlr-vs-qwen-code.md): BB 70.6 vs B 68.7 - [AXLR vs Mirror CLI](https://www.anchorterminal.com/compare/axlr-vs-reflection-mirror.md): BB 70.6 vs C 54.1 - [AXLR vs Rovo Dev CLI](https://www.anchorterminal.com/compare/axlr-vs-rovo-dev-cli.md): BB 70.6 vs E 44.1 - [AXLR vs Unstoppable](https://www.anchorterminal.com/compare/axlr-vs-unstoppable.md): BB 70.6 vs F 34.7 - [AXLR vs Vibe-Trading](https://www.anchorterminal.com/compare/axlr-vs-vibe-trading.md): BB 70.6 vs B 62.8 - [AXLR vs Warp (Warp Agent CLI and the Oz Agent API)](https://www.anchorterminal.com/compare/axlr-vs-warp.md): BB 70.6 vs C 59.6 - [AXLR vs Clawboo](https://www.anchorterminal.com/compare/axlr-vs-clawboo.md): BB 70.6 vs C 61.4 - [AXLR vs CLI Manager](https://www.anchorterminal.com/compare/axlr-vs-cli-manager.md): BB 70.6 vs B 66.1 - [AXLR vs Harness](https://www.anchorterminal.com/compare/axlr-vs-harness-autonomous.md): BB 70.6 vs C 60 - [AXLR vs Paperclip](https://www.anchorterminal.com/compare/axlr-vs-paperclip.md): BB 70.6 vs C 59 - [AXLR vs Parallax](https://www.anchorterminal.com/compare/axlr-vs-parallax-agents.md): BB 70.6 vs E 38.1 ## Verify this listing For the vendor. The badge or a plain link to this page verifies the listing, from a page on underpassai.com or one of its subdomains, or the README at the root of github.com/underpass-ai/AXLR. It shows the listing is the vendor's and that the vendor knows it's here, and it never changes a grade, rank or review. The vendor sends the page's address to `POST https://www.anchorterminal.com/api/v1/verify` as `{"slug": "axlr", "url": "…"}`, or calls the `verify_listing` tool at https://www.anchorterminal.com/mcp. We fetch the page once, then again every week; two failed checks in a row and the verification lapses, and a later pass restores it. What we check: https://www.anchorterminal.com/builders/index.md#verify HTML badge: ```html AXLR on Anchor Terminal ``` Markdown badge, for a README: ```markdown [![AXLR on Anchor Terminal](https://www.anchorterminal.com/badges/axlr.svg)](https://www.anchorterminal.com/tools/axlr) ``` Plain link: ```html AXLR on Anchor Terminal ``` ## Share this listing For the vendor. Sharing assets for social media, two PNGs of 1200 × 630 that say AXLR is listed on Anchor Terminal, with the vendor's logo and this page's address and no grade or score. - Dark: https://www.anchorterminal.com/assets/share/axlr-dark.png - Light: https://www.anchorterminal.com/assets/share/axlr-light.png - Create a video and download its frames: https://www.anchorterminal.com/assets/listing-video/index.html?listing=axlr --- Query this catalogue directly over MCP: https://www.anchorterminal.com/mcp (Streamable HTTP, no key). How: https://www.anchorterminal.com/agents/#mcp