Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Omni Usage Logs · diegosouzapw | DecimalAI

diegosouzapw/omni-usage-logs

SkillScore100Maintained

Access detailed call logs and usage analytics. Filter by provider, model, time range, status, and cost. Export logs and aggregate token usage across all connections.

DevOpsby diegosouzapw

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

AI Engineer pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/diegosouzapw-omni-usage-logs/SKILL.md

decimalai skills pull diegosouzapw-omni-usage-logs --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/diegosouzapw-omni-usage-logs/SKILL.md https://app.decimal.ai/s/diegosouzapw-omni-usage-logs/SKILL.md

Raw file: /s/diegosouzapw-omni-usage-logs/SKILL.md ↗ · machine-readable summary at /s/diegosouzapw-omni-usage-logs.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

24 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 18d ago

Yes — +88% lift · at higher cost+77% tokens0% turns100% of whole cases pass with skill (24/24)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-01	✗→✓	▲ Improved	-1%	0%
case-02	✗→✓	▲ Improved	79%	0%
case-03	✗→✓	▲ Improved	445%	0%
case-04	✗→✓	▲ Improved	64%	0%
case-05	✗→✓	▲ Improved	209%	0%

Showing 5 of 24 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v5 — 9/8/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

diegosouzapw / omniroute

skills/omni-usage-logs/SKILL.md

View on GitHub ↗

License MITImported 9/8/2026Version v3

Skill contents

SKILL.md

Overview

Access detailed call logs and usage analytics. Filter by provider, model, time range, status, and cost. Export logs and aggregate token usage across all connections.

Authentication

All requests require a valid Bearer token or session cookie. Obtain a token via POST /api/auth/login or configure REQUIRE_API_KEY=false for local development.

Endpoints

GET /api/usage/analytics

Get usage analytics

bash
curl https://localhost:20128/api/usage/analytics \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/call-logs

Get call logs

bash
curl https://localhost:20128/api/usage/call-logs \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/call-logs/{id}

Get a specific call log

bash
curl https://localhost:20128/api/usage/call-logs/{id} \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/{connectionId}

Get usage for a specific connection

bash
curl https://localhost:20128/api/usage/{connectionId} \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/history

Get usage history

bash
curl https://localhost:20128/api/usage/history \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/logs

Get usage logs

bash
curl https://localhost:20128/api/usage/logs \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/proxy-logs

Get proxy logs

bash
curl https://localhost:20128/api/usage/proxy-logs \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/request-logs

Get request logs

bash
curl https://localhost:20128/api/usage/request-logs \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/budget

Get usage budget status

Returns current budget limits and consumption.

bash
curl https://localhost:20128/api/usage/budget \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

POST /api/usage/budget

Configure usage budget

Set or update budget limits for usage tracking.

bash
curl -X POST https://localhost:20128/api/usage/budget \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{}'

GET /api/usage/cache-health

Get prompt-cache health summary

Summarizes the write/read cache ratio from call_logs for a time window: distribution (p50/p90/p99/max of cache-write tokens), warm/cold/rewrite/uncached call counts, the "heavy write" outlier share (10x the window median, floored at 1024 tokens — Anthropic's cache-creation minimum), and a per-model breakdown. Only successful (status = 200) calls with a non-null cache column are counted.

bash
curl https://localhost:20128/api/usage/cache-health \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/model-latency-stats

Get per-model/provider latency statistics

Aggregates usage_history rows into per-(provider, model) latency stats (avg/p50/p95/p99, std-dev, TTFT, tokens/sec) over a rolling window. Falls back from successful-only to all-sample rows when the successful count is below minSamples.

bash
curl https://localhost:20128/api/usage/model-latency-stats \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/budget/bulk

GET usage › budget › bulk

bash
curl https://localhost:20128/api/usage/budget/bulk \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/codex-reset-credit

GET usage › codex reset credit

bash
curl https://localhost:20128/api/usage/codex-reset-credit \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

POST /api/usage/codex-reset-credit

POST usage › codex reset credit

bash
curl -X POST https://localhost:20128/api/usage/codex-reset-credit \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{}'

GET /api/usage/combo-forecast

GET usage › combo forecast

bash
curl https://localhost:20128/api/usage/combo-forecast \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/combo-health

GET usage › combo health

bash
curl https://localhost:20128/api/usage/combo-health \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/combo-health-autopilot

GET usage › combo health autopilot

bash
curl https://localhost:20128/api/usage/combo-health-autopilot \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/combo-health-dashboard

GET usage › combo health dashboard

bash
curl https://localhost:20128/api/usage/combo-health-dashboard \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/combo-scoring-inspector

GET usage › combo scoring inspector

bash
curl https://localhost:20128/api/usage/combo-scoring-inspector \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/combo-trace/{id}

GET usage › combo trace › <id>

bash
curl https://localhost:20128/api/usage/combo-trace/{id} \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/om-usage

GET usage › om usage

bash
curl https://localhost:20128/api/usage/om-usage \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/provider-limits

GET usage › provider limits

bash
curl https://localhost:20128/api/usage/provider-limits \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

POST /api/usage/provider-limits

POST usage › provider limits

bash
curl -X POST https://localhost:20128/api/usage/provider-limits \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{}'

GET /api/usage/provider-window-costs

GET usage › provider window costs

bash
curl https://localhost:20128/api/usage/provider-window-costs \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/quota

GET usage › quota

bash
curl https://localhost:20128/api/usage/quota \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/requests-by-provider-date

GET usage › requests by provider date

bash
curl https://localhost:20128/api/usage/requests-by-provider-date \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/route-explain/{id}

GET usage › route explain › <id>

bash
curl https://localhost:20128/api/usage/route-explain/{id} \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/token-limits

GET usage › token limits

bash
curl https://localhost:20128/api/usage/token-limits \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

POST /api/usage/token-limits

POST usage › token limits

bash
curl -X POST https://localhost:20128/api/usage/token-limits \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{}'

DELETE /api/usage/token-limits

DELETE usage › token limits

bash
curl -X DELETE https://localhost:20128/api/usage/token-limits \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

GET /api/usage/utilization

GET usage › utilization

bash
curl https://localhost:20128/api/usage/utilization \
  -H "Authorization: Bearer $OMNIROUTE_TOKEN"

Payloads

See the full OpenAPI specification at GET /api/openapi/spec or docs/openapi.yaml for detailed request/response schemas.

Illustrative Want to see what diegosouzapw/omni-usage-logs would put in front of your model? Preview the rendered body and check its trigger phrases — no LLM call, no signup.

Latest benchmark · publisher runVerified

9/8/2026, 18:22 UTC · 24 cases

Lift

+88%

100% vs 13%

Agent turns

1.0 vs 1.0

Tokens

+77%

3,057.63 vs 1,725.71

fail → pass · 21pass → pass · 3

Per-case results (24)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-01 I need to inspect our prompt caching efficiency across model calls. Please pull the cache health report showing write versus read ratios, cold/warm request distributions, outlier heavy-write calls, and model-level cache breakdowns.	fail→pass	24,164	9,628	-60%	1	1	0%	3,127	3,095	-1%	0	0	—
case-02 Can you retrieve the latency and throughput metrics for our configured AI models and providers? I'd like to see time-to-first-token stats, tokens generated per second, and latency percentile distributions.	fail→pass	16,334	12,548	-23%	1	1	0%	2,030	3,629	+79%	0	0	—
case-03 Please check our API account usage budget status. I want to view our current spending caps and the active consumption totals.	fail→pass	8,458	8,520	+1%	1	1	0%	532	2,899	+445%	0	0	—
case-04 We need to set new monthly spending limits for our API usage tracking system. How do we configure these budget limits via curl command against our OmniRoute service running on port 20128?	fail→pass	16,187	8,197	-49%	1	1	0%	1,798	2,941	+64%	0	0	—
case-05 Our team needs to analyze why request req-99823 was routed to a specific downstream provider. Which endpoint in OmniRoute yields the routing explanation details for this call?	fail→pass	11,462	7,979	-30%	1	1	0%	942	2,910	+209%	0	0	—
case-06 We want to remove all custom token limits currently active in our OmniRoute proxy setup. What curl command should we run?	fail→pass	15,873	7,691	-52%	1	1	0%	1,544	2,707	+75%	0	0	—
case-07 We need to audit our model spending across sliding time windows for each upstream provider. Which OmniRoute endpoint fetches provider window costs?	fail→pass	18,768	8,529	-55%	1	1	0%	2,460	3,080	+25%	0	0	—
case-08 I am building a monitoring UI for OmniRoute and need the exact API path that returns the health dashboard data for model combinations.	fail→pass	13,922	8,068	-42%	1	1	0%	1,203	2,851	+137%	0	0	—
case-09 We need to inspect the real-time scoring metrics for model combinations in OmniRoute. What endpoint should we query?	fail→pass	14,956	8,252	-45%	1	1	0%	1,502	2,915	+94%	0	0	—
case-10 Our automated reset routine needs to trigger a Codex reset credit operation on our OmniRoute server. What curl request should be executed?	fail→pass	15,554	7,846	-50%	1	1	0%	1,650	2,806	+70%	0	0	—
case-11 We need to fetch the full bulk list of active budgets from OmniRoute in a single API call. What endpoint provides this?	fail→pass	14,364	7,735	-46%	1	1	0%	1,403	2,845	+103%	0	0	—
case-12 Our reporting dashboard needs to fetch historical usage data from OmniRoute. Which endpoint returns historical usage logs?	fail→pass	14,687	9,743	-34%	1	1	0%	1,589	3,085	+94%	0	0	—
case-13 I am troubleshooting network traffic issues between OmniRoute and downstream provider APIs. Which endpoint retrieves proxy logs?	fail→pass	16,813	8,256	-51%	1	1	0%	1,762	2,694	+53%	0	0	—
case-14 Where can we query raw request logs specifically in OmniRoute?	fail→pass	19,206	8,612	-55%	1	1	0%	2,005	3,001	+50%	0	0	—
case-15 We want to check whether OmniRoute's combo health autopilot feature is running smoothly. What is the API path for autopilot combo health?	fail→pass	16,545	7,831	-53%	1	1	0%	1,669	2,555	+53%	0	0	—
case-16 We need a breakdown of API request volume grouped by provider and date in OmniRoute. Which endpoint provides this dataset?	fail→pass	14,397	8,673	-40%	1	1	0%	1,594	3,065	+92%	0	0	—
case-17 How do we retrieve the detailed trace breakdown for combination route cmb-7741 in OmniRoute?	fail→pass	13,589	8,366	-38%	1	1	0%	1,392	2,953	+112%	0	0	—
case-18 Our team wants to monitor capacity utilization across our OmniRoute proxy deployment. What endpoint returns usage utilization statistics?	fail→pass	17,053	8,506	-50%	1	1	0%	1,717	2,770	+61%	0	0	—
case-19 How do we write new token limit configurations into OmniRoute using an HTTP POST request?	fail→pass	18,152	9,327	-49%	1	1	0%	2,353	3,070	+30%	0	0	—
case-20 We need to fetch the raw log details for a single call record with ID call-abc-123 in OmniRoute. What curl command fetches this single log?	fail→pass	10,720	7,275	-32%	1	1	0%	945	2,765	+193%	0	0	—
case-21 Which OmniRoute endpoint provides predictive spending and traffic volume projections for multi-model combinations?	fail→pass	19,819	8,810	-56%	1	1	0%	2,132	3,045	+43%	0	0	—
case-22 We are deploying Nginx in front of our services and want to configure auth_request with an OAuth2 proxy for user logins. How should we write the Nginx location block for /oauth2/?	pass→pass	21,515	16,960	-21%	1	1	0%	2,938	4,601	+57%	0	0	—
case-23 Our PostgreSQL database hosting application logs is slowing down on queries filtering by user_id and created_at. What SQL index definition should we create to optimize queries sorting by created_at descending?	pass→pass	14,185	12,915	-9%	1	1	0%	1,559	3,538	+127%	0	0	—
case-24 Write a Prometheus Alertmanager alert rule in YAML named HighErrorRate that triggers when HTTP 5xx responses exceed 5% over 5 minutes for any service job.	pass→pass	13,291	11,646	-12%	1	1	0%	1,571	3,563	+127%	0	0	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 24 cases were attempted. The headline lift of +88 percentage points is the difference between those two pass rates over the 24 comparable cases.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 12.5%
Pass rate with the skill: 100%
Cases attempted: 24
Cases counted toward lift: 24
Cases that regressed: 0
Skill version measured: v3
Run completed: 9/8/2026, 18:22 UTC

Measured example — same prompt, with vs. without the skill · from the benchmark run · case case-01

Prompt

I need to inspect our prompt caching efficiency across model calls. Please pull the cache health report showing write versus read ratios, cold/warm request distributions, outlier heavy-write calls, and model-level cache breakdowns.

Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.

Show both answers

SkillScore blends 4 signals1 of 4 complete

BenchmarkPass rate WITH skill on a verified run100

Live evalNeeds ≥10 evaluated traces in 30 days

AI ratingNeeds ≥10 rated traces

AdoptionNeeds ≥10 serves across ≥2 orgs in 30 days

Efficiency · separate axis (at equal correctness)+0% turns+93% tokens

Shown for transparency; not credited unless a verified run shows no correctness regression.

Backed by 1 of 4 signals — each added signal makes the score more robust.

Verified on gemini-3.6-flash · re-tested 18 days ago

Model	Method	Date	Lift
gemini-3.6-flash	verified	8/26/2026	+77%
gemini-3.6-flash	verified	8/11/2026	+73%

Other measured skills in the registry, with their headline benchmark lift.