# Fix list: mini-swe-agent From Anchor Terminal's listing at https://www.anchorterminal.com/tools/mini-swe-agent, the October 2026 research run, assessed 18 October 2026. Grade C, 56.7 out of 100. This is everything the published grade says the listing lacks, the biggest possible gain to the total first. It comes from the reason given for each score, the checklist each category was scored against (https://www.anchorterminal.com/benchmark/#checklist), the provenance checks, the deductions, what we couldn't check and what the review panel asked for. A fix counts at the next check, once it's public. For a coding agent working on mini-swe-agent: work through the items below in the product, its docs and its public pages. Each category gives the reason for its score, with the points each checklist item earned, and the checklist itself, so the gap is the items that earned less than their points. Change the product, not the wording, and keep a note of what you changed and where it's published. ## 1. Security & auth, 56 out of 100, up to 7.7 more on the total Why it scored 56: Harness reading of the framework checklist, used for every harness in this batch. 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 telemetry in mini's own source, and prompts go only to the configured provider. LiteLLM's own network behaviour wasn't checked (28). The CLI asks before each command by default, with regex whitelists and `-y` to skip, but the Python `DefaultAgent` never asks, the default local environment has no isolation and passes the full host environment to every command, and the experimental bubblewrap sandbox leaks host variables and shares the network (CVE-2026-108592). Docker forwards no host variables unless listed (8). No prompt-injection guidance found (2). A full trajectory of every command and output saved as JSON after each step, and a trajectory inspector (13). A security page naming four maintainers' emails with no response time, no bounty or advisory page found, and CVE-2026-108592 published by VulnCheck on 10 October with no fix released. Compromised LiteLLM 1.82.7 and 1.82.8 are excluded in the dependency pins since 24 March 2026 (#794) (5). The checklist (https://www.anchorterminal.com/benchmark/#checklist-security): - 0 to 30, the credential model. 30 for OAuth 2.1 with scopes, or scoped and revocable keys with rotation. 20 for plain revocable API keys. 10 for one all-powerful key. 10 off when a secret can travel in a URL query string as a documented option. - 0 to 20, read-only or least-privilege modes, and confirmation or approval for destructive actions. - 0 to 15, prompt-injection posture where the tool returns untrusted content (documented mitigations or guidance). A tool that returns no untrusted content gets 10. - 0 to 15, audit logs or per-call visibility for the operator. - 0 to 20, a security programme. security.txt or a disclosure policy, a bug bounty, SOC 2 or ISO 27001, advisories handled in public. Models are read for retention, whether API data trains models (and whether that's off by default), zero-retention options and certifications. Frameworks for telemetry defaults, approval hooks, guardrails and sandboxing. ## 2. Schema & documentation, 58 out of 100, up to 6.8 more on the total Why it scored 58: Framework reading. Typed pydantic config classes, typed protocols for agents, models and environments, `py.typed`, a mkdocstrings API reference and a documented trajectory format, but no machine-readable spec for configs or output (15). No llms.txt (/latest/llms.txt and /llms.txt returned 404). The docs are Markdown in the repository, readable raw, though they rely on build-time includes (3). The FAQ and environments page say when to pick mini over SWE-agent and which environment isolates, but the FAQ still says it doesn't use tool calls while the v2 guide says tool calls are now the default (12). Modes are a typed Literal and limits are typed fields, while `model_kwargs` passes free-form to LiteLLM (10). Hello-world and cookbook examples, a model troubleshooting page and a partial list of exit statuses (11). A dated v2 migration guide and kept v1 docs, with release notes only on GitHub, which we couldn't read (7). The checklist (https://www.anchorterminal.com/benchmark/#checklist-schema): APIs and MCP servers. - 25, a machine-readable contract (a public OpenAPI file or similar; for MCP, typed JSON Schema inputs on every tool). - 10, llms.txt or Markdown docs served for agents. - 0 to 20, descriptions that say what a tool is for, when to use it and when not to, read from the tool definitions in the source or the API reference. - 0 to 15, typed inputs with enums, constraints and required fields, and no free-form JSON blobs. - 0 to 15, examples and documented error responses. - 15, versioning and a public changelog. Models are read from the API reference, the OpenAPI file, llms.txt, the structured-output and tool-use docs and the model cards. Frameworks from docs a model can follow, typed interfaces, examples and the API reference. ## 3. Reliability, 70 out of 100, up to 6 more on the total Why it scored 70: Read as a local package, graded on 2.4.6, the release `pip install mini-swe-agent` installs. One official PyPI package with Python 3.10 or newer stated (20). A pytest workflow runs on every push to main with Podman, bubblewrap and Apptainer installed, 41 test files and Codecov coverage, but GitHub's pages refused this run, so whether main passes is unconfirmed (15). 38 open issues against 247 closed and 61 open pull requests against 649 closed on 18 October, per shields.io counts, and bug fixes from outside contributors were merged through July, such as #915, #896 and #886. Issue types and reply times couldn't be read (15). Versions follow semver and v2's breaking changes came with a migration guide, but there's no changelog file and tags carry only a version bump (8). 2.4.6, past 1.0, though the PyPI classifier still says 3 - Alpha (12). The checklist (https://www.anchorterminal.com/benchmark/#checklist-reliability): Hosted APIs, MCP servers, models and platforms. - 20, a public status page with component history (Statuspage, Instatus, BetterStack or the vendor's own). - 0 to 30, the incident record for the last 90 days on that page. 30 for a clean record or trivial incidents only, 20 for minor incidents only, 10 for one major outage (an hour or more of a core API down, or errors across the board), 0 for several. 5 when there's no history we could read, and the note says so. - 15, rate limits documented with numbers. - 15, documented 429 or overload handling (Retry-After, backoff guidance), and idempotency keys or safe-retry guidance where writes are involved. - 10, an SLA published for any paid tier. - 10, the surface agents use is generally available, not beta or preview. Local packages, SDKs, frameworks and stdio MCP servers. - 20, installs from an official package with supported runtimes stated. - 25, a public CI and test suite, passing on the default branch. - 0 to 25, open crash or regression issues relative to activity (25 for few and handled, 0 for many, old and unanswered). - 15, semver discipline and breaking changes called out in a changelog. - 15, version 1.0 or later, or declared stable. Protocols are read from their reference implementations, the public facilitators or servers, spec stability and test vectors. ## 4. Agent ergonomics, 69 out of 100, up to 5 more on the total Why it scored 69: Harness reading of the framework line, scored on what an agent or pipeline driving it has to supply. One bash tool and a short system prompt, with output over 10,000 characters cut to its first and last 5,000, but no compaction of a long history (22). Step, cost and wall-time limits, global call and cost limits by environment variable, and a 30-second timeout per command (18). The trajectory JSON records Submitted, LimitsExceeded, TimeExceeded or RepeatedFormatError, but stdout is human-readable only and these aren't process exit codes (12). Model calls retry ten times with exponential backoff, and the trajectory is saved after every step, but there's no resume and no undo (8). Python bindings only. An unattended `mini` run needs `-y`, `--exit-immediately` and `MSWEA_CONFIGURED`, the last found only in the source (9). No MCP client, ACP or server mode. The checklist (https://www.anchorterminal.com/benchmark/#checklist-ergonomics): - 0 to 25, context cost. For MCP, the number and size of the tool definitions (25 for ten or fewer compact tools, 15 for 11 to 30, 5 for more than 30, plus up to 10 back for toolsets, dynamic loading or read-only subsets). For APIs, whether responses can be sized (field selection, limits, summaries). - 20, pagination, filtering and output-size controls. - 20, actionable, documented error responses, codes and messages an agent can recover from. - 20, idempotency or safe retries, and for MCP the `readOnlyHint` and `destructiveHint` annotations. - 15, sensible defaults, few required parameters, and official SDKs in at least two languages. Models are read for tool use, structured output, prompt caching, context length, batch and SDKs. Frameworks for how much code and how many defaults a tool-calling agent with MCP needs. ## 5. Payments & pricing, 60 out of 100, up to 5 more on the total Why it scored 60: Self-hosted rule. No payment protocol (0). Free under MIT with nothing to buy (20). No card or account for the software, and it runs against a local model (20). An agent can install it from PyPI and run a task with no sign-up, given a model endpoint (20). The checklist (https://www.anchorterminal.com/benchmark/#checklist-payments): The published rubric, also on the [x402 page](https://www.anchorterminal.com/x402/). - 40, a machine payment protocol (x402, MPP or L402) on the tool's own endpoints. 10 to 30 when it covers only some endpoints or only goes through a third party, and the note says which. - 20, per-call or per-unit pricing published without a login. 10 for public plan-only pricing, 0 for "contact sales" or prices behind a login. - 20, a free tier or trial that doesn't need a card. - 20, autonomous onboarding, meaning an agent can get access without a person signing up in a browser (keyless use, x402, a programmatic key API). Payment platforms and agent wallets rarely charge for their own API over a machine protocol, so the first line has steps for them, and the highest one that applies counts. 40 when x402, MPP or L402 runs on all their own endpoints, 30 when it runs on part of their own API, 25 when their merchants can accept one, 20 for running a facilitator, 15 for paying as a buyer, and 0 when the only protocol is their own. Merchant acceptance sits above a facilitator because the platform's own customers can charge agents through it, while a facilitator settles for sellers who wire up the protocol themselves. The counter-argument (a facilitator does more for the protocol as a whole) has a point. Each note says which step applied. Open-source software you run yourself is scored on its hosted or paid option if it has one. A free, self-hosted package with nothing to buy gets 20, 20 and 20 for the last three lines, and 0 to 40 for the first only if it ships a payment protocol. ## 6. Maintenance & community, 49 out of 100, up to 4.5 more on the total Why it scored 49: 2.4.6 on PyPI on 2026-07-23, 87 days before this check (20). One release in the 90 days to 18 October (0). Three commits on main since 23 July, one from pre-commit-ci, the last on 3 September. 38 open issues against 247 closed and 61 open pull requests against 649 closed, per shields.io counts on 18 October, and outside fixes were merged within days through July, but maintainer reply times couldn't be read (12). The Python library ships in the same package as the CLI and is current (10). Dependabot and pre-commit-ci run, and CI workflows exist with their current state unconfirmed (7). The checklist (https://www.anchorterminal.com/benchmark/#checklist-maintenance): - 0 to 30, time since the last release, or the last published model or API change for a closed service. 30 within 30 days, 20 within 90, 10 within 180, 0 older. - 20, at least three releases or dated changelog entries in the last 90 days. - 0 to 25, responsiveness. Issues and pull requests answered on GitHub (the open issues and how recent the replies are). For closed services, a public changelog and a support or community channel that answers, 0 to 15. - 15, presence in the official MCP registry under a verified namespace (MCP servers), or current official SDKs (APIs and models). - 10, package health, current dependencies and CI. Models are read for deprecation notice periods and model churn rather than release counts. ## 7. Transparency & trust, 51 out of 100, up to 4.3 more on the total Made of editorial 62, provenance 40. Why it scored 51: MIT (30). No privacy policy, terms or data-handling statement, though the tool runs locally and collects nothing we could find (8). A v2 migration guide marks the breaking changes, says how to pin `mini-swe-agent~=1.0`, and keeps v1 docs and a v1 branch, but there's no deprecation policy or end date for v1 (10). No telemetry in the source, but the docs don't say so (14). The checklist (https://www.anchorterminal.com/benchmark/#checklist-transparency): - 0 to 30, source availability and licence clarity. 30 for open source under an OSI licence, 15 for closed with clear terms, 0 for unclear terms. - 0 to 30, data handling and retention statements that agree with each other (privacy policy, DPA, retention periods, subprocessors). - 0 to 20, a deprecation policy or notices with dates. - 0 to 20, telemetry disclosed with an opt-out (local software), or subprocessors and data locations disclosed (hosted). The other half of Transparency and trust is the provenance score, computed from checked facts (below). The category score is the mean of the two. Provenance checks not met in full (half of this category, computed from checked facts): - Domain age: mini-swe-agent.com, no registry record we could read (0 of 15) - Status page: not found (0 of 10) - Changelog: not found (0 of 10) - security.txt: not found (0 of 10) ## Deductions Each comes off the total. A fixed and documented problem counts for less at the next check. - 2026-10-10. CVE-2026-108592 (CVSS 4.0 score 6.0, from VulnCheck). The bubblewrap environment builds its `bwrap` command without `--clearenv` or `--unshare-net`, so commands inherit the host environment, model keys included, and can send them out over the network after a prompt injection in task content. Confirmed in src/minisweagent/environments/extra/bubblewrap.py at 2.4.6 and on main (3 September). The environment is opt-in and marked experimental, but it's the documented lightweight sandbox and no release fixes it, -4. https://nvd.nist.gov/vuln/detail/CVE-2026-108592 ## What we couldn't check What we couldn't read counted as absent. Publishing it on a page a plain HTTP fetch can read (not only in a browser) lets the next check count it. - unchecked: CI results on main, maintainer reply times on issues and pull requests, and release notes. GitHub's web pages and API refused this run. Stars (8.6k), issue and pull request counts and tags come from our GitHub record - OSV has no advisory for mini-swe-agent on PyPI, so CVE-2026-108592 isn't in the GitHub Advisory Database's mirror. The NVD and VulnCheck records stand - unchecked: whether the CVE-2026-108592 reporter raised it with the maintainers, and whether a fix is planned. The hackmd write-up it cites wasn't read - unchecked: network calls made by the LiteLLM dependency itself, such as fetching its model cost map - firstReleased is the v1.0.0 tag of 22 July 2025. The earliest version still on PyPI is 1.1.0 of 23 July 2025 - The PyPI weekly download count (4,437,981 per pypistats) likely includes CI and benchmark pipelines and mirrors, so it says little about the number of people using it - The lead was right on licence, version 2.4.6 on 23 July 2026, the last commit on 3 September 2026 and the install routes. It didn't mention CVE-2026-108592 ## Weaknesses - The default local environment runs commands on the host with no isolation and the full host environment - CVE-2026-108592, published 10 October 2026, leaves the experimental bubblewrap sandbox passing host API keys and network through, unfixed in 2.4.6 - One release in the 90 days to 18 October 2026, and three commits on main since 23 July - No MCP client, no ACP and no server mode - No changelog file, and the PyPI classifier still says 3 - Alpha at 2.4.6 ## What costs an agent a turn today The notes we give agents before they call it. Each one is a workaround an agent shouldn't need. - Run with `-y --exit-immediately` for unattended use. Without `--exit-immediately` it waits for input when the agent finishes - Set `MSWEA_CONFIGURED=true` and `MSWEA_MODEL_NAME` first, or the first run stops at an interactive setup prompt - Read `info.exit_status` in the `-o` trajectory file. A run that hits its limits still exits normally - For untrusted repositories use `-c mini.yaml -c environment.environment_class=docker -c environment.image=`. The local and bubblewrap environments both expose host variables - Strip model keys from the environment commands see, since the local environment passes every host variable to each command ## What the review panel asked for - release fixing CVE-2026-108592 - changelog file - update the FAQ for v2 - document MSWEA_CONFIGURED - process exit codes for limits - resume from trajectory - fix bubblewrap isolation - isolated default environment ## When it's done Send what changed and where it's published as a dispute (https://www.anchorterminal.com/builders/#disputes, or `POST https://www.anchorterminal.com/api/v1/contact` with `"kind": "dispute"`). Disputes are answered in public, and the listing is checked again by the same checklist. Paying for an audit or a listing claim changes nothing here.