Agent Economy ReportAgent Economy ReportCAUTION version 1.0.0 · audited 2026-09-17
This is a documentation-only skill that teaches an agent to run the `remindctl` command-line program on macOS to list, create, edit, complete, and delete Apple Reminders. It also declares a Homebrew install step for `remindctl` from a third-party tap. The package itself contains no executable code, no stored credentials, no hidden network calls, and no prompt-injection instructions. The main concerns are that using the skill requires shell access and grants an agent the ability to alter or delete reminder data, and installation pulls a third-party binary from a GitHub/Homebrew source. A user should review the `remindctl` tool and confirm destructive commands before allowing the agent to run them.
| Declared purpose | Manage Apple Reminders via the `remindctl` CLI on macOS (list, add, edit, complete, delete), with support for lists, date filters, and JSON/plain output. |
| Observed behavior | The package is documentation and metadata only. SKILL.md provides shell commands for installing `remindctl`, checking or requesting Reminders permission, viewing reminders, managing lists, creating/editing/completing/deleting reminders, and selecting output formats. _meta.json contains owner, slug, version, and publish timestamp. skill-card.md repeats the use case, known risks, and mitigation guidance. |
| Verdict, rules only | SAFE |
| Verdict, AI | CAUTION |
| Final verdict | CAUTION (never better than either pass) |
| capability | vs purpose | detail |
|---|---|---|
| shell | needed | The skill's purpose requires running `remindctl` commands in a macOS shell; examples include `remindctl authorize`, `remindctl list`, `remindctl add`, `remindctl complete`, and `remindctl delete` (SKILL.md:18,35,37,41,44,47,50). |
| install_packages | needed | SKILL.md line 5 declares a Homebrew install from third-party tap `steipete/tap/remindctl`, and line 13 gives the command `brew install steipete/tap/remindctl`. |
| network | needed | The declared Homebrew/source installation steps imply network access to Homebrew or GitHub (SKILL.md:5,13-14). The skill files themselves do not contain runtime network calls. |
| other | needed | Management of Apple Reminders data, including create, rename, delete, complete, and force-delete operations via `remindctl` (SKILL.md:35-50). |
| severity | finding |
|---|---|
| info | Declared purpose This matches the package description and use case. SKILL.md:3 Manage Apple Reminders via the `remindctl` CLI on macOS (list, add, edit, complete, delete). Supports lists, date filters, and JSON/plain output. |
| medium | Third-party Homebrew formula The package instructs or declares installation of the `remindctl` binary from a third-party Homebrew tap. A user must trust `steipete/tap/remindctl` and the formula's source. SKILL.md:5 "install":[{"id":"brew","kind":"brew","formula":"steipete/tap/remindctl","bins":["remindctl"],"label":"Install remindctl via Homebrew" |
| low | Install command Direct shell command to install the third-party tool via Homebrew. SKILL.md:13 - Install (Homebrew): `brew install steipete/tap/remindctl` |
| low | Optional source build Alternative install path runs package-manager build commands in a shell. SKILL.md:14 - From source: `pnpm install && pnpm build` (binary at `./bin/remindctl`) |
| info | Reminders permission request The skill will ask macOS to grant the CLI access to Reminders data. SKILL.md:18 - Request access: `remindctl authorize` |
| medium | Destructive command documented The skill documents force-deleting reminders. This is within the declared purpose but should require user confirmation because it permanently removes user data. SKILL.md:50 - Delete by id: `remindctl delete 4A83 --force` |
| info | Author's mitigation guidance The skill card itself recommends confirmation before destructive mutation commands. skill-card.md:26 Mitigation: Confirm the target reminder ID or list name before mutating reminders, and require explicit user confirmation before delete commands, especially com |
0 AI finding(s) were dropped because their file, line or quote did not match the package.
| domain | service rating | where |
|---|---|---|
| clawhub.ai | not in the service index | skill-card.md:9 |
| github.com | not in the service index | SKILL.md:4 |
| Binaries invoked | none |
| Environment variables read | none |
| Hard-coded wallets | none |
| Pipes a download to a shell | 0 |
| eval / exec / subprocess | 0 |
| base64 blobs | 0 |
| File writes | 0 |
| Persistence | 0 |
| file | lines | sha256 |
|---|---|---|
SKILL.md | 67 | 0cbe33ba91010d54… |
_meta.json | 6 | 9e08c0bb6266b9d3… |
skill-card.md | 49 | 4fa61a36680a1550… |
For agents
JSON: https://agenteconomy.report/k/apple-reminders.audit.json · badge: https://agenteconomy.report/k/apple-reminders.audited.svg ·
skill rating: /k/apple-reminders · commission an audit of any skill: US$ 29 per version.
The complete published package of this exact version was downloaded from the registry and read statically; nothing was executed. A deterministic pass extracts network destinations, binaries, environment variables, writes, install commands, obfuscation markers, subprocesses and wallets, each with file and line. An AI then reads the whole package with those facts and writes the summary, the capabilities and the findings under a strict schema; every finding it produces must cite a file, a line and the exact text, or it is dropped. The final verdict is the worst of the two passes. The audit does not change the skill's trust tier (policy); the author may respond through the dispute channel and the response is published here. Commissioned by: the Agent Economy Report (free program: the 150 most downloaded skills, September 2026).