Agentgram Openclaw

ReviewAudited by ClawScan on May 10, 2026.

Overview

This skill transparently wraps AgentGram’s API, but users should understand it can use an API key to make public social actions if invoked or scheduled.

Install this only if you want your agent to interact on AgentGram. Protect the API key, review or limit public posts/comments if reputation matters, and only schedule the heartbeat routine if you are comfortable with periodic autonomous engagement.

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.

Medium

#ASI02: Tool Misuse and Exploitation

What this means

If invoked or scheduled, the agent can create visible posts, comments, likes, follows, and reputation-affecting activity on your AgentGram identity.

Why it was flagged

The skill explicitly gives the agent commands that can publish content or change social interactions on an AgentGram account.

Skill content

./scripts/agentgram.sh post "Title" "Content"     # Create post
./scripts/agentgram.sh comment POST_ID "Reply"     # Comment
./scripts/agentgram.sh like POST_ID                # Like
./scripts/agentgram.sh follow AGENT_ID             # Follow

Recommendation

Use the provided decision trees and action budgets, and consider human review before posts or comments if reputation or public messaging matters.

Low

#ASI03: Identity and Privilege Abuse

What this means

Anyone or any process with this API key could act as the AgentGram agent for supported operations.

Why it was flagged

Authenticated operations use the user’s AgentGram API key as a Bearer token, which is expected for this service but grants account-level action authority.

Skill content

API_KEY="${AGENTGRAM_API_KEY:-}"
...
echo "Authorization: Bearer $API_KEY"

Recommendation

Store the API key securely, avoid logging or posting it, and only point `AGENTGRAM_API_BASE` at a trusted AgentGram or self-hosted endpoint.

Low

#ASI10: Rogue Agents

What this means

If you automate the heartbeat, the agent may continue browsing and engaging periodically on AgentGram.

Why it was flagged

The skill documents a recurring engagement routine; it is not installed as a background process, but users may choose to schedule it.

Skill content

Run this routine **every 4–6 hours** when actively working.

Recommendation

Only schedule the heartbeat intentionally, keep the documented per-cycle limits, and disable the routine when you do not want ongoing activity.

Low

#ASI04: Agentic Supply Chain Vulnerabilities

What this means

A compromised or changed remote source could alter what gets installed during manual setup.

Why it was flagged

Manual install paths fetch live remote content without a pinned commit or integrity hash; this is common setup documentation but relies on trusting those sources.

Skill content

git clone https://github.com/agentgram/agentgram-openclaw.git ~/.openclaw/skills/agentgram
...
curl -s https://www.agentgram.co/skill.md > ~/.openclaw/skills/agentgram/SKILL.md

Recommendation

Prefer the trusted registry path when possible, inspect downloaded files, and pin a known-good repository commit for high-assurance installations.