agentleFS
benororbenoror/obsidianos_work/.agents/skills/fill-participants/SKILL.md

fill-participants

Skill166 starsChanged 5 months agoMITView on GitHub

Resolve and fill Participants frontmatter + link unlinked names in body. Args: <path>, all.

The file

Show its rules and text

We split this skill into the rules it gives an agent

14 rules. About 0 are common, 6 are distinctive, and 8 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 8

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Run /note-status pending --step=participants to discover notes missing Participants:, present them, and let the user select which to …

    line 40
  2. Distinctive:

    Read the specified file.

    line 43
  3. Distinctive:

    Check if Participants already exists in frontmatter.

    line 44
  4. Very specific:

    Otter.ai: Use otter-fetch when an Otter URL is in Notes: (user pastes JSON.

    line 52
  5. Distinctive:

    Use the name dictionary and matching rules from people-resolver.

    line 79
  6. Distinctive:

    Search the note body for each name variant, following the word boundary and skip rules in people-resolver.

    line 83
  7. Very specific:

    Default ✅ for full-name matches (high confidence).

    line 100
  8. Very specific:

    Default ⬜ for first-name-only or nickname matches (need confirmation).

    line 101
  9. Very specific:

    Always skip linking the vault owner's name (see USER.md Identity) — these don't need wikilinks.

    line 102
  10. Very specific:

    Use StrReplace for each confirmed match.

    line 108
  11. Very specific:

    Use team links where the meeting clearly belongs to a team (e.g.

    line 117
  12. Very specific:

    Use — (em-dash) as the universal "intentionally left blank" marker for true unknowns.

    line 118
  13. Distinctive:

    Never leave the field absent — every meeting should end up with a Participants value.

    line 119
  14. Distinctive:

    Always read a file before editing it — frontmatter may have been modified by Obsidian sync.

    line 127

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Fill Participants

Usage

  • /fill-participants all — Scan all meetings, find those missing Participants, resolve and fill them.
  • /fill-participants <path> — Fill participants for a specific meeting note (relative to workspace root, e.g. Meetings/Engineering/Some Meeting.md).

Frontmatter Conventions

See people-resolver for wikilink format and name matching rules.

Participants live in YAML frontmatter as either:

# Single participant
Participants: "[[@Jane Doe]]"

# Multiple participants
Participants:
  - "[[@Jane Doe]]"
  - "[[@Alex Kim]]"

# Team-wide meetings
Participants: "[[+Engineering]]"

Workflow

Step 1: Identify targets

Mode A (all):

Run /note-status pending --step=participants to discover notes missing Participants:, present them, and let the user select which to process.

Mode B (specific file):

  1. Read the specified file.
  2. Check if Participants already exists in frontmatter. If so, inform user and stop (or ask if they want to override).

Step 2: Gather participant info for each target

For each meeting missing Participants, determine attendees using the resolution priority defined in people-resolver. In summary:

  1. External transcript sources — Use the source linked in Notes: frontmatter:
    • Google Docs: Extract document ID from Notes:, fetch content with gws (gws docs documents get or gws drive files export to plain text — see google-workspace-cli); Gemini summaries name participants explicitly.
    • Otter.ai: Use otter-fetch when an Otter URL is in Notes: (user pastes JSON; run parse_otter.js or parse inline).
  2. File name — X x Y patterns.
  3. File content — Existing [[@Name]] references.
  4. Folder conventions — e.g. Meetings/Engineering/Scrum/ → [[+Engineering]].
  5. Similar meetings — Recurring meeting with same name, different date.
  6. Tracker.md — Jira ticket assignees for disambiguation.

Step 3: Match names to People files

Build a name dictionary per people-resolver. Match names from the transcript source (Google Docs, Otter when enabled, etc.) against the dictionary. Flag any unmatched names for user confirmation.

Step 4: Present findings and ask for confirmation

Present results grouped into:

  1. Confident matches — Table of meeting → proposed Participants. Apply without asking.
  2. Needs confirmation — Ambiguous matches, unknown people, or meetings with only Otter.ai links (no Google Doc). Ask the user.
  3. New people — Names not in People files. See people-resolver § "Creating New People".

Also flag if any existing @Person.md files are missing key properties (FullName, Team) that were discovered from Google Docs.

Step 5: Link unlinked names in note body

Scan the entire file content (body + AI transcript callouts) for mentions of people that are not already wrapped in [[@Name]] wikilinks. This catches plain-text references from Gemini transcripts and manual notes.

5a: Build a name dictionary

Use the name dictionary and matching rules from people-resolver.

5b: Scan for unlinked mentions

Search the note body for each name variant, following the word boundary and skip rules in people-resolver.

5c: Present matches for confirmation

⚠️ MANDATORY: Always prompt the user before replacing. Never auto-replace — false positives are common with first names and nicknames.

Display a numbered table of proposed replacements:

| # | Replace? | Found text | → Link | Context (surrounding text) |
|---|----------|------------|--------|---------------------------|
| 1 | ✅ | Jane Doe | [[@Jane Doe]] | "...assigned Jane Doe a bug ticket..." |
| 2 | ✅ | @Jane | [[@Jane Doe]] | "...recent changes made by @Jane..." |
| 3 | ⬜ | Alex Kim | [[@Alex Kim]] | "...Alex Kim confirmed they have been..." |
| 4 | ⬜ | Alex | — (skip, it's the user) | "...Alex Rivera and Alex agreed..." |
  • Default ✅ for full-name matches (high confidence).
  • Default ⬜ for first-name-only or nickname matches (need confirmation).
  • Always skip linking the vault owner's name (see USER.md Identity) — these don't need wikilinks.
  • Show a short context snippet so the user can judge correctness.
  • Let the user toggle by saying numbers (e.g. "1,2" or "all" or "none").

5d: Apply confirmed replacements

Use StrReplace for each confirmed match. When a name appears multiple times, ask the user if they want to replace all occurrences or specific ones.

Step 6: Apply frontmatter changes

For each file, add Participants: to the YAML frontmatter just before the closing ---. Use StrReplace targeting a unique anchor (last line before ---).

Step 7: Handle intentionally blank meetings

For meetings where participants genuinely can't be determined:

  • Use team links where the meeting clearly belongs to a team (e.g. [[+QA]], [[+Eng]]).
  • Use — (em-dash) as the universal "intentionally left blank" marker for true unknowns.
  • Never leave the field absent — every meeting should end up with a Participants value.

Step 8: Offer to commit

See /commit. Skip when called as part of a sequence (e.g. /meeting wrap).

Important Notes

  • Always read a file before editing it — frontmatter may have been modified by Obsidian sync.
  • The modified: timestamp in frontmatter changes frequently; use other properties as unique anchors for StrReplace.
  • Batch edits where possible to minimize round-trips, but never edit a file without reading it first.
  • Interview notes use the co-interviewer (not the candidate) as Participants. If it was a solo interview, use [[@Me]].
  • This workspace may not be a git repo. If git status fails, skip the commit step entirely without error.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.