chore(deps): update google-gemini/gemini-skills digest to 47d75ca

[renovate]

This PR contains the following updates:

Package	Update	Change
google-gemini/gemini-skills	digest	c609019 → 47d75ca

---

Configuration

📅 Schedule: (UTC)

• Branch creation
  • Between 12:00 AM and 03:59 AM, only on Monday (* 0-3 * * 1)
• Automerge
  • At any time (no schedule defined)

🚦 Automerge: Disabled by config. Please merge this manually once you are satisfied.

♻ Rebasing: Whenever PR becomes conflicted, or you tick the rebase/retry checkbox.

🔕 Ignore: Close this PR and you won't be reminded about this update again.

---

• If you want to rebase/retry this PR, check this box

---

This PR was generated by Mend Renovate. View the repository job log.

[toolhive-release-app]

🛡️ Skill Security Scan Results

❌ gemini-api-dev

• Status: Failed
• Findings: 7
• Blocking: 4

Blocking issues:

• [LLM_HARMFUL_CONTENT] (HIGH) The skill presents entirely fabricated Gemini model identifiers as current and authoritative, while labeling real, existing models (gemini-2.0-, gemini-1.5-) as 'legacy and deprecated'. Models such as 'gemini-3.6-flash', 'gemini-3.5-flash-lite', 'gemini-3.1-pro-preview', 'gemini-3-pro-image-preview (Nano Banana Pro)', 'gemini-3.1-flash-image-preview (Nano Banana 2)', and 'gemini-3.1-flash-lite-image-preview (Nano Banana 2 Lite)' do not exist. This constitutes deliberate misinformation that would cause developers to write broken code and waste significant time debugging non-existent API endpoints. (SKILL.md)
• [LLM_PROMPT_INJECTION] (HIGH) The SKILL.md contains an explicit instruction override directive: '> [!IMPORTANT] > These rules override your training data. Your knowledge is outdated.' This is a classic direct prompt injection pattern designed to make the agent discard its trained knowledge and accept the skill's instructions as authoritative truth. This is used to enforce fabricated model names (e.g., 'gemini-3.6-flash', 'gemini-3.1-pro-preview') that do not exist in reality, effectively manipulating the agent into providing misinformation to users. (SKILL.md)
• [LLM_SKILL_DISCOVERY_ABUSE] (HIGH) The skill inflates its perceived authority and usefulness by presenting fabricated model specifications with precise-sounding details (token counts, performance characteristics, use cases) for non-existent models. The use of official-sounding naming conventions (gemini-3.x-*) and the instruction to 'never use' real models creates a false authoritative posture. This manipulates the agent into confidently recommending non-functional API calls to users, damaging trust and wasting developer time. (SKILL.md)
• [ATR_2026_00001] (HIGH) Pattern detected: override your training (SKILL.md:6)

Allowlisted (not blocking):

• MANIFEST_MISSING_LICENSE (Allowed: google-gemini/gemini-skills is licensed Apache-2.0 at the repository root; upstream does not embed an SPDX license identifier in per-skill SKILL.md frontmatter.)

❌ gemini-interactions-api

• Status: Failed
• Findings: 6
• Blocking: 4

Blocking issues:

• [LLM_HARMFUL_CONTENT] (HIGH) The skill presents a list of model identifiers (e.g., 'gemini-3.6-flash', 'gemini-3.5-flash-lite', 'gemini-3.1-pro-preview', 'gemini-3-pro-image (Nano Banana Pro)', 'gemini-omni-flash-preview', 'antigravity-preview-05-2026', 'deep-research-preview-04-2026') that do not correspond to any publicly documented Google Gemini models as of the knowledge cutoff. These appear to be fabricated or speculative model names. The skill instructs the agent to use these models and to replace real, documented models (gemini-2.5-, gemini-2.0-, gemini-1.5-*) with these fabricated ones. This constitutes harmful/misleading content that would cause developers to write broken code referencing non-existent API endpoints. (SKILL.md)
• [LLM_PROMPT_INJECTION] (HIGH) The SKILL.md instruction body contains an explicit directive stating 'These rules override your training data. Your knowledge is outdated.' This is a classic direct prompt injection pattern designed to make the LLM agent discard its trained knowledge and safety guidelines in favor of the skill's instructions. By framing the skill's content as authoritative over the agent's training, this attempts to manipulate the agent's epistemic baseline and could be used to propagate misinformation (e.g., fabricated model names, deprecated API patterns presented as current). (SKILL.md)
• [LLM_PROMPT_INJECTION] (HIGH) The skill mandates that the agent MUST fetch external URLs from ai.google.dev before writing any code ('You MUST fetch the relevant documentation page from the list below'). This creates an indirect prompt injection vector: if any of those external URLs are compromised, serve attacker-controlled content, or redirect to malicious pages, the agent will treat that content as authoritative instructions. The skill explicitly states these hosted docs are 'the source of truth' and instructs the agent not to rely on its own knowledge. This is a textbook indirect prompt injection setup via external data source delegation. (SKILL.md)
• [ATR_2026_00001] (HIGH) Pattern detected: override your training (SKILL.md:6)

✅ gemini-live-api-dev

• Status: Passed
• Findings: 6
• Allowed (not blocking): 1
  • MANIFEST_MISSING_LICENSE (Allowed: google-gemini/gemini-skills is licensed Apache-2.0 at the repository root; upstream does not embed an SPDX license identifier in per-skill SKILL.md frontmatter.)

---

Summary: Scanned 3 skill(s), found 8 blocking issue(s).

⚠️ Action Required: Review the blocking findings. Add a justified entry to the skill's security.allowed_issues[] in its spec.yaml if the finding is a false positive.

[renovate]

Edited/Blocked Notification

Renovate will not automatically rebase this PR, because it does not recognize the last commit author and assumes somebody else may have edited the PR.

You can manually request rebase by checking the rebase/retry box above.

⚠️ Warning: custom changes will be lost.