Clarify YouTube transcript wrapper usage

This commit is contained in:
2026-08-27 19:14:11 -04:00
parent a72c5d5545
commit dad835fff7
2 changed files with 4 additions and 6 deletions

View File

@@ -27,3 +27,5 @@ Personal agent skills packaged through Nix.
Treat deliberate ambiguity as part of the skill's design. Treat deliberate ambiguity as part of the skill's design.
Customize only the parts required to fit the local workflow. Customize only the parts required to fit the local workflow.
Do not turn incidental choices into new contracts. Do not turn incidental choices into new contracts.
- Document helper commands with their location, not only a bare `./helper` invocation.
Agents may run skill instructions from the caller's current working directory instead of the skill directory.

View File

@@ -11,11 +11,7 @@ Create durable output when the user requests it or active project conventions re
## Helper command ## Helper command
Run from this skill directory: Run the `fetch_transcript` wrapper provided by this skill for the YouTube URL.
```bash
./fetch_transcript "<youtube_url>"
```
The command prints JSON to stdout: The command prints JSON to stdout:
@@ -31,7 +27,7 @@ The command exits non-zero on failure and writes the error to stderr.
## Ingest workflow ## Ingest workflow
1. Run `./fetch_transcript` for the YouTube URL. 1. Run the `fetch_transcript` wrapper provided by this skill for the YouTube URL.
2. Read the returned `title` and `transcript`. 2. Read the returned `title` and `transcript`.
3. Build source understanding from the transcript. 3. Build source understanding from the transcript.
4. Respond to the user's request in chat by default. 4. Respond to the user's request in chat by default.