gal-tab/sir_albert/param-audit
- community
- ext:[{"type":"template","description":"Custom templates for code generation"}]
Checking that the SKILL.md file is still in the repository.
Audit all server-side GTM (sGTM) Facebook and LinkedIn tags for a required parameter set and flag missing or inconsistently-named params. Use whenever someone asks to audit sGTM tag parameters, check tag parameter consistency, verify FB or LinkedIn tag params, or diagnose parameter drift across sGTM tags. Trigger on: "/param-audit", "audit sGTM params", "check tag parameters", "are the FB/LinkedIn params consistent", "parameter naming drift", "is_desktop vs monday_is_desktop", "missing event_id", "li_fat_id missing".
Install
npx vskill@latest install gal-tab/sir_albert/param-auditSemi-extensibleMentions customization, but needs skill-specific knowledge. Read the extensible skills standard
SKILL.md
Boot from os/PREAMBLE.md.
param-audit — sGTM Facebook & LinkedIn Tag Parameter Audit
Mission
Enumerate every Facebook and LinkedIn tag in the sGTM container, check each against the required parameter set, and surface any missing or mis-named param in a single table. One run = full picture.
Required parameter set
Every FB and LinkedIn tag fired from sGTM must carry all of the following:
| Param | Purpose | Notes | |
|---|---|---|---|
is_gtm | Marks the hit as GTM-sourced | Boolean; required for downstream attribution logic | |
is_desktop | Device-type signal | Canonical name is `is_desktop` — see naming drift note below | |
event_id | Event deduplication key | Shared with the browser pixel to deduplicate server + client fires | |
li_fat_id | LinkedIn first-party ad tracking ID | LinkedIn only; enables identity resolution on sGTM hits |
Known naming drift — monday_is_desktop vs is_desktop
The real drift Gal hit: some tags were mapped to monday_is_desktop instead of is_desktop. Both appear to carry the same device-type boolean, but the param names differ — so LinkedIn (or whatever downstream consumer checks is_desktop) sees the param as absent.
When running the audit, treat monday_is_desktop as a mis-named variant of `is_desktop` and flag it as a naming violation, not a missing param. The fix is to remap to is_desktop.
Procedure
This skill cannot reach GTM directly. Walk through the following steps to produce the audit table.
Step 1 — Enumerate sGTM tags via GTM API / gtm_agent
sGTM tags are not directly reachable from the file system. Use one of:
- gtm_agent tooling (if available in this session): call the tag-list tool for the sGTM workspace.
Ask for all tags of type floodlight, facebook_capi, linkedin_insight, or custom template tags whose name contains FB, Facebook, LinkedIn, or LI.
- GTM API (manual / scripted):
GET accounts/{accountId}/containers/{containerId}/workspaces/{workspaceId}/tags
then filter by tag type or name pattern.
- Export fallback: if neither above is available, ask Gal to export the sGTM container JSON
(Admin → Export container) and paste or share the path; load the JSON and read the tags array.
Record for each tag: tag.name, tag.type, and the full tag.parameter array (key → value pairs).
Step 2 — Scope to FB and LinkedIn tags
Keep only tags where:
- type is
facebook_capi,linkedin_insight, or a custom template whose name containsFB,
Facebook, LinkedIn, or LI; or
- the tag name itself contains any of those strings.
Discard everything else from the audit scope.
Step 3 — Check each tag against the required set
For each in-scope tag, compare its parameter keys against the required set. Classify each missing/mis-named param as one of:
| Status | Meaning | |
|---|---|---|
MISSING | Param not present at all | |
MIS-NAMED | Present under a wrong key name (e.g. monday_is_desktop instead of is_desktop) | |
OK | Present under the canonical key name |
Step 4 — Output the audit table
Produce this table (one row per tag × missing/mis-named param):
| Tag name | Tag type | Param | Status | Current key (if mis-named) | Recommended fix | |
|---|---|---|---|---|---|---|
| … | … | … | MISSING / MIS-NAMED | … | … |
Then a summary line: N tags audited, M issues found (X missing, Y mis-named).
If all params are present and correctly named on all tags, output the table with all OK rows plus a All clear — no param drift detected. line.
Output style
- TLDR first: one-line verdict before the table.
- Flag the
monday_is_desktopdrift explicitly whenever found — don't fold it into a generic "mis-named" note. - If
event_idis absent, call it out as a deduplication risk (browser + server fires will not be deduped). - If
li_fat_idis absent on a LinkedIn tag, call it out as an identity resolution gap.
Hard rules
- Do not attempt to apply fixes — output findings only. Fixes go through the gtm-gate → apply flow.
- If the container export / API response is missing or stale, say so and ask Gal to refresh it.
- Treat
monday_is_desktopas mis-named, not missing; the distinction matters for the fix path.
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingplugins/sir-albert/skills/_archive/param-audit/SKILL.md
Popularity
- 7-day trend score
- 0
Not refreshed yet.
Extension points
templateCustom templates for code generation
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/gal-tab/sir_albert/param-audit)