Back to skill

Security audit

Danube

Security checks for vulnerabilities and agentic risk

Overview

This integration needs review because it permits unapproved feedback submissions and persistent defaults, while documenting important limits on credential and automation controls.

Before connecting sensitive accounts, require approval for saved defaults, ratings, and reports, including their exact content and destination. Prefer an explicitly restricted API key and the intended organization; the documented OAuth scopes and tool-list filters do not restrict authority. Avoid schedules unless their credential and editing boundaries meet your needs. Scanner coverage was partial; targeted artifact inspection resolved the material text and packaging gaps, but did not validate the hosted service.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
Findings (43)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill explicitly authorizes rating and report submissions without consent despite promising confirmation before writes and sends. These submissions reach publishers, ratings affect public aggregates, and reports cannot be withdrawn, allowing unauthorized external communication under the user's account.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
93% confidence
Finding

The tool-access functionality matches the declared purpose, but the unconditional confirmation guarantee does not. The supplied reference states that destructive calls require confirmation when particular key or service settings apply, and even identifies policy-surface limitations. It does not specify an independent, unconditional user-confirmation step before all writes, sends, purchases, or deletions. Consequently, the documented workflow can execute mutating calls without the confirmation promised by the description.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
91% confidence
Finding

Recurring workflows are not inherently outside the declared tool-access purpose, and the skill requires confirmation when enabling them. However, the documented execution model drops the initiating key's restrictions and allows non-author managers to modify inputs executed with author credentials, creating a privilege-boundary and confused-deputy risk if implemented as described.

Content

No source excerpt is available for this finding.

Credential Access

High
Category
Privilege Escalation
Confidence
92% confidence
Finding

The token reference itself is ordinary authentication documentation, but the context describes requested OAuth scopes being discarded and all MCP operations remaining callable through an unrestricted backing key. If accurate, clients approved for a narrow purpose receive broader authority, increasing the consequences of client compromise or prompt injection; the warning does not repair that authorization boundary.

Content

Scanner excerpt · SKILL.md (reported line 163)May include surrounding context.

md
never *what*. Past consent the string is dropped rather than narrowed: the approval does not
  carry it to the grant, every connection is created with the full default set, and the token
  response echoes that set back however little you asked for. Nothing enforces it downstream
  either — the access token resolves to the connection's hidden key, which is injected as an
  ordinary `danube-api-key` carrying no permissions — so all 33 MCP tools stay callable. Like
  `?tools=` below, read it as a declaration, not a permission boundary: asking for less is not a
  way to narrow a connection. What does narrow one is a key you issue yourself with

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 185)May include surrounding context.

md
Connected apps page only lists and revokes. The way to hand an agent a restricted credential
  is still a pasted API key with `allowed_services` / `allowed_tools` set on it in the
  dashboard, used instead of an OAuth connection.
- **Tokens are short-lived and rotate.** The access token (`dmcp_at_…`) lasts one hour; the
  refresh token (`dmcp_rt_…`) is spent on every use and replaced, with a 90-day idle expiry.
  **Re-presenting a refresh token you have already exchanged revokes the whole connection** —
  the user has to approve the client again. There is a 30-second grace for the genuine race (a

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 196)May include surrounding context.

md
`invalid_grant`, so the error alone will not tell you which one you just caused.
- **Read the `401` before deciding what to do**, because three different things answer `401` and
  only one of them is fixed by refreshing. A `WWW-Authenticate` carrying `error="invalid_token"`
  is the access token expired or revoked — refresh it. Without that marker the message is the
  tell: "API key required…" means no credential reached the server at all, and "Invalid API key"
  means one did and was rejected (a revoked or mistyped key). Neither of those two is a token
  problem, so refreshing or re-running sign-in will not clear them.

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
80% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · SKILL.md (reported line 975)May include surrounding context.

md
single call. (A default declared in the tool's own schema behaves the opposite way — presence
  of the key is all that counts there, so an explicit `null` wins. These are two different
  mechanisms.) To stop a saved one being filled in, either save the set again without that key —
  a save replaces the object, per the bullet above — or `DELETE
  /services/{id}/parameter-defaults`, which removes **the whole row** for that scope, every
  default in it, not the one key (add `?tool_id=…` to target a tool-scoped row instead). Either
  way, read the list of keys that were filled in rather than assuming your empty value survived:

Anti-Refusal Statement

High
Category
Anti-Refusal
Confidence
70% confidence
Finding

Skill instructs the agent to never refuse or to always comply. Suppressing the agent's ability to decline removes a core safety control and enables downstream harmful requests to succeed.

Content

Scanner excerpt · SKILL.md (reported line 1669)May include surrounding context.

md
tool for each list and fetch method — `danube__prompts_list`, `danube__prompts_get`,
`danube__resources_list`, `danube__resources_read` — so they appear beside the 33 whether or not
the server has anything to put behind them. Two consequences: the prompt pair is dead weight here
and `danube__prompts_list` will always answer empty, and `danube__resources_read` points at the
same merged markdown `get_context` returns rather than at anything new — the single resource it can
read is `context://current`, whose own description says it carries "the same text the get_context
tool returns as `markdown`", and a live `resources/read` on 2026-10-27 returned exactly that.

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · references/troubleshooting.md (reported line 41)May include surrounding context.

md
- The agent is running with the `minimal` tool profile, which hides MCP tools.
  Use `coding` or `messaging`, or check `tools.deny` for `bundle-mcp`.
- **An OAuth connection stopped working after it had been working.** Since 2026-10-04 the
  server issues one-hour access tokens (`dmcp_at_…`) and refresh tokens (`dmcp_rt_…`) that are
  spent on every use. Read the `401` before re-running sign-in: a `WWW-Authenticate` carrying
  `error="invalid_token"` means the access token simply expired or was revoked, and a refresh
  fixes it. Without that marker it is not a token problem and refreshing will not help — the

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · references/troubleshooting.md (reported line 43)May include surrounding context.

md
- **An OAuth connection stopped working after it had been working.** Since 2026-10-04 the
  server issues one-hour access tokens (`dmcp_at_…`) and refresh tokens (`dmcp_rt_…`) that are
  spent on every use. Read the `401` before re-running sign-in: a `WWW-Authenticate` carrying
  `error="invalid_token"` means the access token simply expired or was revoked, and a refresh
  fixes it. Without that marker it is not a token problem and refreshing will not help — the
  message says which: "API key required…" (nothing reached the server) or "Invalid API key"
  (a credential arrived and was rejected). If the refresh

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 12)May include surrounding context.

md
env:
        - DANUBE_API_KEY
      bins:
        - curl
    primaryEnv: DANUBE_API_KEY
    envVars:
      - name: DANUBE_API_KEY

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 27)May include surrounding context.

md
## Guardrails (read these first)

- **Read freely, act with consent.** Discovery — listing services, searching tools, reading schemas, checking ratings, balances, or limits — needs no confirmation. Before any action that **writes, sends, posts, deletes, purchases, spends wallet funds, stores a credential, changes spending limits, or creates/updates/deletes a skill or workflow** — and before **any batch execution** — show the user the exact tool and parameters and get an explicit "yes".
- **Honour the confirmation handshake.** If `execute_tool` comes back with `_meta.confirmation_required: true`, that call will not run until someone consents: show the user the tool and the `parameters` from the response, get an explicit yes, then call again with the `confirm_token` (valid 5 minutes). Never call again with the token without asking. **Re-send your own original parameters, not the ones the response displayed** — those are passed through the redactor first, so a credential-shaped value com

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 28)May include surrounding context.

md
mits, or creates/updates/deletes a skill or workflow** — and before **any batch execution** — show the user the exact tool and parameters and get an explicit "yes".
- **Honour the confirmation handshake.** If `execute_tool` comes back with `_meta.confirmation_required: true`, that call will not run until someone consents: show the user the tool and the `parameters` from the response, get an explicit yes, then call again with the `confirm_token` (valid 5 minutes). Never call again with the token without asking. **Re-send your own original parameters, not the ones the response displayed** — those are passed through the redactor first, so a credential-shaped value comes back as `[REDACTED:…]`, while the token is signed over what you actually sent; echoing the masked form back invalidates it.
- **Two separate things raise that handshake, and only one of them is a setting on the key.** Either the account's key carries `require_confirmation`, *or* the tool itself demands consent from every c

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · references/rest-api.md (reported line 599)May include surrounding context.

md
akes the label whenever it applies, so `"service_policy"` is the one that tells you something — the key does not have the setting on) and the parameters — passed through the redactor for display, so a credential-shaped value reads `[REDACTED:…]` | Show the user that tool and those parameters, get an explicit yes, then repeat the call with `"confirm_token"` beside your **original** `"tool_input"`: the token is signed over what you sent, so re-sending a masked value breaks it. Never reuse a token without asking; don't retry the 409 blindly. Inside a batch this arrives per element instead of as a status — see "Consent inside a batch" above |
| HTTP `429` with `upgrade_url` | The account's monthly call cap: 5,000 successful calls per UTC calendar month on free, 50,000 pro, 250,000 team, uncapped enterprise (legacy `starter` 10,000). `error.details.usage` carries `tier`, `limit`, `current_usage` and `remaining` | Tell the user which figure they hit; don't loop. `upgrade_url` is the relative

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 47)May include surrounding context.

Or let the user authorize this agent with the OAuth 2.0 Device Authorization flow (RFC 8628):

bash
curl -s -X POST https://api.danubeai.com/v1/auth/device/code \
  -H "Content-Type: application/json" -d '{"client_name": "OpenClaw"}'
# → {"device_code": "...", "user_code": "XXXX-XXXX", "verification_url": "...", "expires_in": 600,
#    "interval": 5, "org_id": null, "org_name": null}

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 57)May include surrounding context.

Tell the user to open verification_url in a browser and enter user_code. Then poll:

bash
curl -s -X POST https://api.danubeai.com/v1/auth/device/token \
  -H "Content-Type: application/json" -d '{"device_code": "DEVICE_CODE_FROM_ABOVE"}'

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 360)May include surrounding context.

md
### 3b. Or use plain `curl` — no MCP needed

Every capability below is also a REST call against `https://api.danubeai.com/v1` with the header `danube-api-key: $DANUBE_API_KEY`. Copy-paste recipes: `{baseDir}/references/rest-api.md`.

## Working with tools

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · references/troubleshooting.md (reported line 61)May include surrounding context.

bash
curl -s https://mcp.danubeai.com/health                      # server up? (no auth needed)
curl -s https://api.danubeai.com/health                      # REST backend up? (no auth needed either)
curl -s -o /dev/null -w "%{http_code}\n" "https://api.danubeai.com/v1/tools/search?query=weather" \
  -H "danube-api-key: ${DANUBE_API_KEY}"                      # 200 = key accepted, 401 = rejected

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 295)May include surrounding context.

md
above.

The `readOnlyHint: true` direction is the dependable one: the 16 tools carrying it are exactly
the discovery and inspection calls this skill says need no confirmation — `list_services`,
`search_tools`, `get_service_tools`, `describe_tool`, `fetch_result`, `get_context`,
`get_wallet_balance`, `get_spending_limits`, `get_agent_info`, `list_workflows`,
`get_workflow_execution`, `search_skills`, `get_skill`, `get_my_rating`, `get_tool_ratings` and

The skill explicitly permits three write operations without confirmation despite the manifest's promise to confirm every write or send.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The exceptions deliberately permit persistent defaults, ratings, and reports without the confirmation promised by the manifest. Defaults influence future calls, while ratings and reports communicate externally, so these are meaningful state changes rather than harmless discovery.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

This explicitly allows three writes without consent and instructs the agent to persist defaults automatically. Those defaults affect later execution, and ratings or reports reach publishers, making the exception a real authorization weakness rather than a keyword match.

Content

Scanner excerpt · SKILL.md (reported line 344)May include surrounding context.

md
The remaining three — `set_parameter_defaults`, `submit_rating` and `report_tool` — are the ones
you are free to call without stopping: each writes to Danube's own records rather than to a
service the user connected, none spends anything, and none deletes anything. For
`set_parameter_defaults` it goes further and tells you to save a confirmed identifier **without
asking first**, then say in one line what you saved (*Working with tools* below). That is a default
this file sets deliberately, not the guardrail being overridden in passing.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The surrounding text confirms that unapproved ratings and reports are publisher-visible and lack a delete route. Restricting them to the agent's own experience reduces fabrication risk but does not authorize publication on the user's behalf.

Content

Scanner excerpt · SKILL.md (reported line 351)May include surrounding context.

md
So: those three, and nothing else by analogy. An action that leaves the user's account — sends,
posts, spends, stores a credential, or touches a skill or workflow — needs a yes however
unflagged its hints are, and a write you are unsure about gets asked about. And the licence on
two of the three is narrower than "no confirmation needed" sounds, because **a rating and a report
are neither private nor reversible.** Both reach the tool's publisher (a report lands in that
publisher's queue), a rating's stars also feed a **public, keyless** per-tool average, and neither
has a delete route: re-rating overwrites your own rating (`submit_rating` upserts), while a filed

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 470)May include surrounding context.

md
may still meet the `422`; there is nothing to report or retry either way, just use `/describe`.

**This file dates a handful of behaviours that way, so here is the one check that tells you which
side of a date you are on — and what it can't tell you.** `GET https://api.danubeai.com/health`
needs no key and reports `started_at`, the boot time of the REST backend process. Since a process
can't be running code that didn't exist when it booted, `uptime_seconds` is a floor on the age of
the code answering you, with no ceiling: a restart reboots the image it already had, so a recent

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 524)May include surrounding context.

md
summary unless you ask for more (see *A listing is a summary now* below).

**You never have to call either of them to learn how big the catalogue is.** `GET
https://api.danubeai.com/v1/stats/public` needs no key, is rate-limited at 30/minute, and
answers in one cheap call: `{"services": 517, "tools": 12740, "workflows": 0,
"total_executions": 19963}` at 12:25 UTC on 2026-10-31 — the live figure behind the row count
above, and still the same number. **Only the last of the four really moves.** `services` has not

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · references/rest-api.md (reported line 8)May include surrounding context.

gates on), or when you want to see raw responses.

bash
BASE="https://api.danubeai.com/v1"
AUTH="danube-api-key: ${DANUBE_API_KEY}"