OpenTIL
Analysis
OpenTIL appears purpose-aligned for saving and managing TIL entries, but it does use an OpenTIL token and can publish, edit, sync, and delete entries when instructed.
Findings (5)
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.
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.
`publish` | `write:entries` ... `edit` | `read:entries` + `write:entries` ... `delete` | `delete:entries` ... Type "delete" to confirm
The skill can mutate OpenTIL content, including publishing, editing, and permanently deleting entries, but the documented flows include previews or confirmations for high-impact actions.
Agent proactively detects TIL-worthy moments ... Append the suggestion at the end of your normal response ... Capture? (yes/no)
The skill can alter normal agent responses by adding proactive capture suggestions, but it limits suggestions to once per session and requires user acceptance before capture.
Open `{verification_uri}?user_code={user_code}` via `open` (macOS) or `xdg-open` (Linux) ... Use a bash script to poll in a single commandThe auth flow documents local command use to open a browser and poll for authorization, which is purpose-aligned but still involves shell-level actions.
Checks whether tool use, credentials, dependencies, identity, account access, or inter-agent boundaries are broader than the stated purpose.
create a Personal Access Token with `read:entries`, `write:entries`, and `delete:entries` scopes ... `$OPENTIL_TOKEN` ... `~/.til/credentials` file
The skill requires or uses an OpenTIL account token that can read, write, and delete entries, and it can also read tokens from a local profile file.
Checks for exposed credentials, poisoned memory or context, unclear communication boundaries, or sensitive data that could leave the user's control.
All platforms use `~/.til/drafts/` ... Parse the frontmatter ... Read the content body ... POST to API ... On 201 success: delete the local file
The skill stores drafts persistently on disk and later reuses those files during sync, uploading their contents to OpenTIL after the documented sync flow.
