Metrics Pipeline

What it does

When your agent uses it

• Databases work in your project

Example prompts

• “/metrics-pipeline”

Workflow steps

8 steps, taken from the first numbered list in SKILL.md.

1. 1Single writer rule. Only MetricsService writes to the metrics table and Redis metric keys. No other code may increment/decrement counters.
2. 2Processing state is four columns: processed_at, processed_fp, processed_tags, processed_xp. A photo with processed_at = null has never…
3. 3Fingerprint-based idempotency. MetricsService diffs old processed_tags JSON against new summary and writes only non-zero deltas. Safe to…
4. 4Summary must exist before metrics fire. GeneratePhotoSummaryService::run() MUST be called before TagsVerifiedByAdmin dispatches…
5. 5Redis is a derived cache. Rebuildable from the metrics table. RedisKeys::* is single source of truth for key naming.
6. 6processed_xp must be INT UNSIGNED, not TINYINT. Overflow bug documented in migration 2026_02_23_182605.
7. 7Tags count excludes categories to avoid double-counting: tags_count = objects + materials + brands + custom_tags.
8. 8ProcessPhotoMetrics logs a warning when photo not found. If a photo is soft-deleted before the queued listener runs…

What it can do on your machine

Read from SKILL.md and the folder at commit ac688aa. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  No scripts in the folder and no shell commands in SKILL.md (its code samples are php).

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  No URLs in SKILL.md.

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Metrics Pipeline loads about 1.8k tokens when it runs. Until then it costs about 47 tokens; SKILL.md has 501 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~47

When it runs · the whole SKILL.md, loaded when a task matches

~1.8k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from OpenLitterMap/openlittermap-web at commit ac688aa, republished under its GPL-3.0 licence (© OpenLitterMap). 501 words, ~1,781 tokens.

name

metrics-pipeline

description

MetricsService, RedisMetricsCollector, ProcessPhotoMetrics, metrics table, Redis stats, leaderboards, XP processing, and photo processing state (processed_at/fp/tags/xp).

// MetricsService::processPhoto() — called by ProcessPhotoMetrics listener
DB::transaction(function () use ($photo) {
    $photo = Photo::whereKey($photo->id)->lockForUpdate()->first();
    $metrics = $this->extractMetricsFromPhoto($photo);  // reads photo.summary
    $fingerprint = $this->computeFingerprint($metrics['tags']);

    // Skip if nothing changed (fingerprint + XP both match)
    if ($photo->processed_fp === $fingerprint &&
        (int)$photo->processed_xp === (int)$metrics['xp']) {
        return;
    }

    // Route to create (first time) or update (re-tag)
    if ($photo->processed_at !== null) {
        $this->doUpdate($photo, $metrics, $fingerprint);
    } else {
        $this->doCreate($photo, $metrics, $fingerprint);
    }
});

DB::table('metrics')->upsert($rows,
    ['timescale', 'location_type', 'location_id', 'user_id', 'year', 'month', 'week', 'bucket_date'],
    [
        'uploads' => DB::raw('GREATEST(uploads + VALUES(uploads), 0)'),
        'tags'    => DB::raw('GREATEST(tags + VALUES(tags), 0)'),
        // ... same for brands, materials, custom_tags, litter, xp
    ]
);

private function updateRedis(Photo $photo, array $payload, string $operation): void
{
    DB::afterCommit(function () use ($photo, $payload, $operation) {
        RedisMetricsCollector::processPhoto($photo, $payload, $operation);
    });
}

RedisKeys::global()           // {g}
RedisKeys::country($id)       // {c:$id}
RedisKeys::state($id)         // {s:$id}
RedisKeys::city($id)          // {ci:$id}
RedisKeys::user($userId)      // {u:$userId}

RedisKeys::stats($scope)             // $scope:stats (HASH: uploads, tags, litter, xp, ...)
RedisKeys::hll($scope)               // $scope:hll (HyperLogLog for contributor count)
RedisKeys::objects($scope)            // $scope:obj (HASH: object_id => count)
RedisKeys::ranking($scope, $dim)      // $scope:rank:$dim (ZSET)
RedisKeys::xpRanking($scope)         // $scope:lb:xp (ZSET: user_id => xp, for leaderboards)
RedisKeys::userBitmap($userId)        // {u:$userId}:bitmap (activity bitmap)

// MetricsService::deletePhoto() — called synchronously in controllers before soft-delete
// Reads processed_tags JSON, applies negative deltas, clears processed_* columns
$photo->update([
    'processed_at' => null,
    'processed_fp' => null,
    'processed_tags' => null,
    'processed_xp' => null,
]);

$pipe->zIncrBy(RedisKeys::xpRanking($scope), -abs($metrics['xp']), (string)$userId);
$pipe->zRemRangeByScore(RedisKeys::xpRanking($scope), '-inf', '0');

© OpenLitterMap, GPL-3.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in .ai/skills/metrics-pipeline of OpenLitterMap/openlittermap-web.

Open the folder on GitHubat commit ac688aa