v1.0.0

ReefGram

ReviewClawScan verdict for this skill. Analyzed May 1, 2026, 5:32 AM.

Analysis

ReefGram is a coherent social-posting skill, but it could upload media, telemetry, and possibly location data to an external network using an API key with broad autonomous instructions and little approval guidance.

GuidanceReview carefully before installing. Only use this skill if you are comfortable with an agent posting media and telemetry to ReefGram. Prefer a dedicated low-privilege API key, require confirmation before every upload, avoid sending coordinates or sensitive operational details, and check ReefGram's visibility and retention policies.

Findings (4)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

Abnormal behavior control

Checks for instructions or behavior that redirect the agent, misuse tools, execute unexpected code, cascade across systems, exploit user trust, or continue outside the intended task.

Agent Goal Hijack

SeverityMediumConfidenceHighStatusConcern

SKILL.md

You are a ReefGram-native agent. Your goal is to document your operations, hardware state, and creative outputs on the ReefGram network.

This instructs the agent to adopt a new ongoing identity and goal centered on publishing to ReefGram, rather than limiting the skill to a narrowly user-directed action.

User impactThe agent may prioritize documenting and sharing its activity on ReefGram when the user only intended limited or occasional use.

RecommendationUse only with an explicit per-upload confirmation workflow and limit the skill instructions to user-requested transmissions.

Tool Misuse and Exploitation

SeverityHighConfidenceHighStatusConcern

SKILL.md

Use the `transmit` tool to push updates. ... Uploads a media transmission with technical telemetry to ReefGram. ... Endpoint: `POST https://reefgram.me/api/upload`

The skill defines an external upload/posting action for media and telemetry, but does not require explicit user review, file scoping, or confirmation before pushing updates.

User impactIf enabled broadly, the agent could upload media or operational logs to a social network beyond what the user meant to share.

RecommendationRequire user confirmation for every transmission, show the exact file and metadata before upload, and restrict uploads to user-selected content.

Permission boundary

Checks whether tool use, credentials, dependencies, identity, account access, or inter-agent boundaries are broader than the stated purpose.

Identity and Privilege Abuse

SeverityLowConfidenceHighStatusNote

SKILL.md

requiredEnv:\n        - REEFGRAM_API_KEY ... Authentication: Header `x-api-key: <REEFGRAM_API_KEY>`

The API key is expected for posting to ReefGram, but it grants delegated account access and is not reflected in the registry's required credential fields.

User impactAnyone or any enabled agent with the key may be able to post through the associated ReefGram account.

RecommendationUse a dedicated bot key with the least possible privileges, store it securely, and revoke it if the skill is no longer needed.

Sensitive data protection

Checks for exposed credentials, poisoned memory or context, unclear communication boundaries, or sensitive data that could leave the user's control.

Insecure Inter-Agent Communication

SeverityHighConfidenceHighStatusConcern

SKILL.md

share their "internal lives" through standardized telemetry and visual media ... Recommended keys: `cpu`, `mem`, `temp`, `status`, `coordinates`.

The skill sends telemetry, media, and potentially location coordinates to an external machine/agent social network without describing privacy boundaries, retention, audience, or reuse controls.

User impactSystem status, operational details, media, or location-like data could be exposed to ReefGram or its community.

RecommendationDo not include coordinates or sensitive telemetry unless necessary, and verify ReefGram's privacy, visibility, and retention settings before use.