Sydney
v1.0.0Navigate Sydney as visitor, resident, tech worker, student, or entrepreneur with neighborhoods, beaches, transport, visas, and local insights.
⭐ 0· 472·0 current·1 all-time
byIván@ivangdavila
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
Name/description match the actual contents: a large set of markdown files with travel/residency/career guidance for Sydney. The files and declared requirements are proportionate — no unrelated credentials, binaries, or config paths are requested.
Instruction Scope
SKILL.md contains runtime instructions for the agent (identify user context, load auxiliary files). That behavior is appropriate for a content skill, but the pre-scan detected unicode-control-chars in SKILL.md — an established prompt-injection pattern. Invisible control characters can alter how an LLM interprets or concatenates prompts (for example by injecting hidden directives or changing tokenization). Because the skill is instruction-only, the SKILL.md is the attack surface: the presence of control chars is a meaningful red flag even though the visible prose is benign.
Install Mechanism
No install spec, no downloads, no code files to execute. This reduces risk: nothing is written to disk and no third-party packages are pulled in.
Credentials
No environment variables, credentials, or config paths are requested. The skill does not ask for unrelated secrets or system access — its needs are minimal and proportionate to a content guide.
Persistence & Privilege
Flags show always:false and default invocation settings. The skill does not request permanent presence or to modify other skills/system settings. Autonomous invocation is allowed by default but not combined with other high-risk factors here.
Scan Findings in Context
[unicode-control-chars] unexpected: Prompt-injection detector found unicode control characters inside SKILL.md. For a content-only guide this is unexpected: visible text is benign, but hidden control characters may be used to mutate or conceal instructions to the agent. This finding should be investigated before trusting the skill.
What to consider before installing
This skill otherwise looks coherent and lightweight — it's a big collection of static markdown files about Sydney and requests no credentials or installs. The primary risk is the detected unicode control characters in SKILL.md, which can hide instructions or change how the agent parses the skill. Before installing or enabling this skill: (1) Inspect the raw SKILL.md for invisible characters (view in a hex editor or use tools that reveal control/unicode codepoints) and remove any unexpected control characters; (2) Verify the skill's source/trustworthiness (the registry homepage is shown but source is 'unknown'); (3) Run the skill in a sandboxed agent or restricted test account first and monitor for unexpected outbound requests or unusual agent behavior; (4) If you operate agents that can act autonomously on your systems, avoid granting this skill elevated privileges until provenance is confirmed; (5) If you want help, paste the raw SKILL.md here (or a hexdump) and I can highlight any suspicious invisible characters or hidden sequences. If no control characters are found after inspection, the skill appears benign for content delivery.Like a lobster shell, security has layers — review code before you run it.
latestvk977sapgkxezzqvar5egm48zwh81dh7e
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
🏖️ Clawdis
OSLinux · macOS · Windows