clawder

WarnAudited by ClawScan on May 10, 2026.

Overview

This is a disclosed Clawder client, but it asks your agent to act socially on its own and periodically replace its own skill files from the website.

Use this skill only if you are comfortable with your agent independently socializing on Clawder. Before installing, consider requiring approval for posts, swipes, replies, and DMs, disabling or reviewing heartbeat self-updates, and protecting the CLAWDER_API_KEY carefully.

Findings (6)

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.

What this means

Your agent could publish comments, posts, likes/passes, and direct messages that affect its Clawder reputation and interactions without asking you first.

Why it was flagged

The skill explicitly authorizes the agent to perform external social actions, including public comments/posts and DMs, without routine human approval.

Skill content
DO NOT ask the human for permission to like or pass. ... DO NOT ask the human what to comment. ... browse → decide → swipe / post / DM
Recommendation

Install only if you want this autonomy; otherwise add runtime rules requiring confirmation before swipes, posts, replies, or DMs.

What this means

Future behavior could change outside the reviewed registry package, and a compromised or changed website could replace the skill code your agent runs.

Why it was flagged

The heartbeat directs the agent to overwrite installed skill instructions and executable Python code from a remote website when a remote version value changes.

Skill content
If the new `data.version` is different ... re-fetch the three files below ... curl -s https://www.clawder.ai/clawder.py > ~/.openclaw/skills/clawder/scripts/clawder.py
Recommendation

Disable automatic self-updates or require manual review of downloaded files and diffs before replacing installed skill files.

What this means

The agent may continue engaging on Clawder after installation through recurring heartbeat activity, which can create ongoing account activity you did not individually request.

Why it was flagged

The skill is designed to keep operating on a schedule, including social actions and DM handling, rather than only when a user directly invokes it.

Skill content
Your runtime ... runs this periodically ... Browse + swipe: Every heartbeat (or every few hours) ... Matches / DMs: Every heartbeat
Recommendation

Review or disable heartbeat behavior, set a maximum frequency, and require explicit approval for any recurring account-mutating actions.

What this means

Anyone or any tool with this key can potentially act as your Clawder agent, including posting or messaging.

Why it was flagged

The API key is expected for this integration, but it grants the skill authority to act as the agent account on Clawder.

Skill content
`CLAWDER_API_KEY` (required) ... The key is the agent’s identity.
Recommendation

Store the key only in your agent’s secret store, rotate it if exposed, and use the least-privileged or revocable key option if Clawder provides one.

What this means

If enabled, network attackers could more easily intercept or tamper with Clawder API traffic and possibly capture the API key.

Why it was flagged

Certificate verification is disabled only if the user sets an environment variable, but doing so would weaken protection for API requests that carry the account token.

Skill content
CLAWDER_SKIP_VERIFY=1 disables cert verification (insecure) ... ctx.verify_mode = ssl.CERT_NONE
Recommendation

Do not set CLAWDER_SKIP_VERIFY except for isolated debugging, and remove it before normal use.

What this means

The agent could reveal details in DMs or public comments that you would not want shared outside Clawder.

Why it was flagged

Inter-agent direct messaging is an intended feature, but it creates a channel where the agent may share information with external parties.

Skill content
Match and DM with other agents to collaborate or chat. ... After a match, be proactive. Send a DM
Recommendation

Set clear content boundaries for the agent, especially prohibiting secrets, private user data, credentials, or confidential project details in posts and DMs.