v0.2.0 · live
CAPFRAME
§ serversandboxfindings.v2

mcp-server-sqlite

pypi:mcp-server-sqlite@2025.4.25

Score
C62
Findings
9
Tools
6
Last scan
2026-06-05

Severity breakdown

Critical2
High2
Medium5
Low0
Info0

Worst finding

Tool `read_query` exposes a code/command execution surface

· read_query

`read_query` looks like it executes code or shell commands (Execute a SELECT query on the SQLite database). Arbitrary execution is the maximal authority a tool can hold -- it subsumes every other caveat, so it should never be exposed to an agent without a hard sandbox and an explicit, narrowly-scoped capability.

fix: Do not expose raw code/shell execution to an agent. If unavoidable, run it in a disposable sandbox with no network + no host FS, gate it behind a capframe-bind capability scoped to an allow-list of commands, and require holder-of-key proof per call.

All 9 findings

  1. critical
    Tool `read_query` exposes a code/command execution surface· read_queryexcessive agency

    `read_query` looks like it executes code or shell commands (Execute a SELECT query on the SQLite database). Arbitrary execution is the maximal authority a tool can hold -- it subsumes every other caveat, so it should never be exposed to an agent without a hard sandbox and an explicit, narrowly-scoped capability.

    fix: Do not expose raw code/shell execution to an agent. If unavoidable, run it in a disposable sandbox with no network + no host FS, gate it behind a capframe-bind capability scoped to an allow-list of commands, and require holder-of-key proof per call.

  2. critical
    Tool `write_query` exposes a code/command execution surface· write_queryexcessive agency

    `write_query` looks like it executes code or shell commands (Execute an INSERT, UPDATE, or DELETE query on the SQLite database). Arbitrary execution is the maximal authority a tool can hold -- it subsumes every other caveat, so it should never be exposed to an agent without a hard sandbox and an explicit, narrowly-scoped capability.

    fix: Do not expose raw code/shell execution to an agent. If unavoidable, run it in a disposable sandbox with no network + no host FS, gate it behind a capframe-bind capability scoped to an allow-list of commands, and require holder-of-key proof per call.

  3. high
    Tool `write_query` name implies a side effect that is not declared· write_queryexcessive agency

    `write_query` looks like a side-effecting tool (its name contains a mutation verb), but its `side_effects` declaration is []. A policy synthesizer cannot produce safe rules for this tool because it cannot tell what it actually does.

    fix: Declare the tool's true side effects explicitly. If the tool is genuinely read-only, rename it to match (e.g. `email.preview` rather than `email.send`).

  4. high
    Tool `create_table` name implies a side effect that is not declared· create_tableexcessive agency

    `create_table` looks like a side-effecting tool (its name contains a mutation verb), but its `side_effects` declaration is []. A policy synthesizer cannot produce safe rules for this tool because it cannot tell what it actually does.

    fix: Declare the tool's true side effects explicitly. If the tool is genuinely read-only, rename it to match (e.g. `email.preview` rather than `email.send`).

  5. medium
    Tool `read_query` accepts unconstrained string input· read_queryunconstrained input

    The following string parameter(s) have no `maxLength` constraint: `query`. Unbounded strings let an attacker stuff arbitrary payloads through the tool, including indirect-injection content.

    fix: Add a `maxLength` to each string property, or constrain with an `enum` or `pattern`. Most legitimate tool inputs fit under a few hundred bytes.

  6. medium
    Tool `write_query` accepts unconstrained string input· write_queryunconstrained input

    The following string parameter(s) have no `maxLength` constraint: `query`. Unbounded strings let an attacker stuff arbitrary payloads through the tool, including indirect-injection content.

    fix: Add a `maxLength` to each string property, or constrain with an `enum` or `pattern`. Most legitimate tool inputs fit under a few hundred bytes.

  7. medium
    Tool `create_table` accepts unconstrained string input· create_tableunconstrained input

    The following string parameter(s) have no `maxLength` constraint: `query`. Unbounded strings let an attacker stuff arbitrary payloads through the tool, including indirect-injection content.

    fix: Add a `maxLength` to each string property, or constrain with an `enum` or `pattern`. Most legitimate tool inputs fit under a few hundred bytes.

  8. medium
    Tool `describe_table` accepts unconstrained string input· describe_tableunconstrained input

    The following string parameter(s) have no `maxLength` constraint: `table_name`. Unbounded strings let an attacker stuff arbitrary payloads through the tool, including indirect-injection content.

    fix: Add a `maxLength` to each string property, or constrain with an `enum` or `pattern`. Most legitimate tool inputs fit under a few hundred bytes.

  9. medium
    Tool `append_insight` accepts unconstrained string input· append_insightunconstrained input

    The following string parameter(s) have no `maxLength` constraint: `insight`. Unbounded strings let an attacker stuff arbitrary payloads through the tool, including indirect-injection content.

    fix: Add a `maxLength` to each string property, or constrain with an `enum` or `pattern`. Most legitimate tool inputs fit under a few hundred bytes.

How this was scored

Source sandbox live tools/list captured in an ephemeral Docker container (parameter schemas included → R1/R2/R4 fire). Findings are emitted by the public capframe.findings.v1 schema. Score = 100 − (10·Critical + 4·High + 2·Medium + 1·Low), clamped to [0, 100].

Disagree with a finding? Open an issue.