Ready to supercharge your AI agents?

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

Get Started Free →

Send Request Skill · dicklesworthstone | DecimalAI

dicklesworthstone/send-request

NewSkillScoreMaintained

Send HTTP requests with automatic configuration and authentication. Use when testing APIs, webhooks, or any HTTP endpoint.

Testingby dicklesworthstone

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

Automatic

Coding pack

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

pip install decimalai

2 · Get the skill — writes .claude/skills/dicklesworthstone-send-request/SKILL.md

decimalai skills pull dicklesworthstone-send-request --out .claude/skills/

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

curl -fsSL --create-dirs -o .claude/skills/dicklesworthstone-send-request/SKILL.md https://app.decimal.ai/s/dicklesworthstone-send-request/SKILL.md

Raw file: /s/dicklesworthstone-send-request/SKILL.md ↗ · machine-readable summary at /s/dicklesworthstone-send-request.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

23 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 2h ago

—+21% tokens0% turns83% of whole cases pass with skill (19/23)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-04	✗→✓	▲ Improved	25%	0%
case-05	✗→✓	▲ Improved	42%	0%
case-06	✗→✓	▲ Improved	16%	0%
case-07	✗→✓	▲ Improved	27%	0%
case-09	✗→✓	▲ Improved	3%	0%

Showing 5 of 23 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/10/2026

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

dicklesworthstone / pi_agent_rust

tests/ext_conformance/artifacts/npm/pi-super-curl/skills/send-request/SKILL.md

View on GitHub ↗

License NOASSERTIONImported 8/10/2026Version v1

Skill contents

SKILL.md

Send Request Skill

Sends HTTP requests with:

Config-based: baseUrl, auth, headers (from .pi-super-curl/config.json)
Named endpoints: Quick access via @endpoint-name
Template variables: {{uuid}}, {{uuidv7}}, {{env.VAR}}, {{timestamp}}
JWT auth: Auto-generates tokens with configurable payload
Env resolution: $VAR syntax for secrets in config

Usage

Run the send-request.cjs script with request parameters:

bash
# First, find the script path (use -L to follow symlinks)
SCRIPT=$(find -L ~/.pi/agent/skills -name "send-request.cjs" 2>/dev/null | head -1)

# Then use it
node "$SCRIPT" <METHOD> "<URL>" [options] 2>&1

Parameters:

METHOD: GET, POST, PUT, PATCH, DELETE
URL: Full URL or @endpoint-name from config
--body '{"key": "value"}': Request body (JSON, supports templates)
--header 'Name: Value': Custom header (repeatable)
--save: Save response to ~/Desktop/api-responses/
--stream: Stream SSE responses

Examples:

bash
# Simple GET request
node "$SCRIPT" GET "https://httpbin.org/get" 2>&1

# POST with JSON body and templates
node "$SCRIPT" POST "@chat" --body '{"id": "{{uuidv7}}", "user": "{{env.USER_ID}}"}' 2>&1

# Named endpoint from config
node "$SCRIPT" GET "@health" 2>&1

# With custom header
node "$SCRIPT" GET "https://api.example.com/data" --header "X-Custom: value" 2>&1

Configuration

The script reads .pi-super-curl/config.json from the current directory (walking up) or home directory:

json
{
  "baseUrl": "$API_BASE_URL",
  "envFile": ".env",
  "auth": {
    "type": "jwt",
    "secret": "$JWT_SECRET",
    "algorithm": "HS256",
    "expiresIn": 3600,
    "payload": {
      "user_id": "{{env.USER_ID}}",
      "role": "authenticated"
    }
  },
  "headers": {
    "Content-Type": "application/json",
    "X-Org-Id": "{{env.ORG_ID}}"
  },
  "endpoints": [
    {
      "name": "health",
      "url": "/health",
      "method": "GET"
    },
    {
      "name": "chat",
      "url": "/api/chat",
      "method": "POST",
      "defaultBody": {
        "chat_id": "{{uuidv7}}",
        "workspace_id": "{{env.WORKSPACE_ID}}"
      }
    }
  ]
}

Template Variables

| Template | Description | |----------|-------------| | {{uuid}}, {{uuidv4}} | Random UUID v4 | | {{uuidv7}} | Time-ordered UUID v7 | | {{timestamp}} | Unix timestamp (seconds) | | {{timestamp_ms}} | Unix timestamp (ms) | | {{date}} | ISO date string | | {{env.VAR}} or {{$VAR}} | Environment variables |

Environment Resolution

Two syntaxes for different contexts:

| Syntax | Use in | Example | |--------|--------|---------| | $VAR | baseUrl, auth.secret, auth.token | "baseUrl": "$API_URL" | | {{env.VAR}} | URLs, headers, body, JWT payload | "user_id": "{{env.USER_ID}}" |

Authentication Types

Bearer Token

json
{"type": "bearer", "token": "$MY_API_TOKEN"}

JWT (auto-generated)

json
{
  "type": "jwt",
  "secret": "$JWT_SECRET",
  "algorithm": "HS256",
  "expiresIn": 3600,
  "payload": {
    "user_id": "{{env.USER_ID}}",
    "email": "{{env.EMAIL}}",
    "role": "authenticated"
  }
}

API Key

json
{"type": "api-key", "token": "$API_KEY", "header": "X-API-Key"}

Basic Auth

json
{"type": "basic", "username": "$USER", "password": "$PASS"}

Output

The script outputs:

[INFO] lines to stderr (method, URL, timing)
Response body to stdout
Raw response saved to /tmp/generation-output.txt (for /scurl-log)
[INFO] Request completed successfully on success
[ERROR] on failure

Want to see if dicklesworthstone/send-request would activate on your phrasing? Try it before you install — no LLM call, no signup.

Latest benchmark · publisher runVerified

8/29/2026, 11:03 UTC · 23 cases

Measured on 20 comparable cases — below the 22 needed to publish a lift, so this result is shown here but not on the registry card or leaderboard.

Lift

+35%

83% vs 48%

Agent turns

0%

1.0 vs 1.0

Tokens

+21%

1,771.7 vs 1,459.3

fail → pass · 9pass → pass · 10fail → fail · 3pass → fail · 1

Per-case results (23)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-01 Can you run a health check against our backend by sending a GET request to the configured @health endpoint using the send-request utility and return the response details?	fail→fail	6,579	5,013	-24%	1	1	0%	894	1,434	+60%	0	0	—
case-02 Please post a message payload to the @chat named endpoint. The body should include a dynamic time-ordered UUID for chat_id and reference the WORKSPACE_ID from environment variables, then provide the returned JSON response.	fail→fail	9,525	7,457	-22%	1	1	0%	2,109	1,768	-16%	0	0	—
case-03 I need to fetch data from https://api.example.com/v1/metrics using a GET request with an additional header 'X-Org-Id: org_123'. Please save the API response to disk and show me the output.	fail→fail	9,630	7,687	-20%	1	1	0%	1,676	1,637	-2%	0	0	—
case-04 I want to test a real-time event stream from https://api.example.com/v1/events using a GET request. Usually I would write a Python `requests` script with `stream=True`, but how should this be executed using our HTTP request helper tool?	fail→pass	10,188	4,079	-60%	1	1	0%	1,566	1,960	+25%	0	0	—
case-05 We want to send a PUT request to https://api.example.com/users/123 with a JSON body containing a random v4 UUID transaction token named `tx_id`. Standard tools often use `$GUID` or `%UUID%`. What exact template tag string should be placed inside the JSON body for the request helper to substitute a v4 UUID?	fail→pass	6,220	3,345	-46%	1	1	0%	1,211	1,717	+42%	0	0	—
case-06 When posting a payload to https://api.example.com/telemetry with a timestamp field, I need the current Unix epoch time in seconds. A colleague suggested `{{timestamp_ms}}`. Which template variable tag specifies Unix timestamp in seconds?	fail→pass	8,646	2,482	-71%	1	1	0%	1,433	1,660	+16%	0	0	—
case-07 We need to send a POST request to https://api.example.com/logs with a JSON body having a `created_at` field set to the current ISO date string automatically. What exact template variable tag does our HTTP tool substitute for ISO dates?	fail→pass	7,506	1,711	-77%	1	1	0%	1,214	1,536	+27%	0	0	—
case-08 In `.pi-super-curl/config.json`, I am setting up JWT auto-generation. Inside the `payload` dictionary, I want the `user_id` claim to load from the `USER_ID` environment variable. Should this claim value be written as `$USER_ID` or `{{env.USER_ID}}`?	pass→pass	19,248	3,758	-80%	1	1	0%	2,320	1,843	-21%	0	0	—
case-09 In `.pi-super-curl/config.json`, we are configuring JWT authentication. For the top-level `secret` field under the `auth` section, which syntax should be used to resolve the secret from the `JWT_SECRET` environment variable?	fail→pass	11,027	2,715	-75%	1	1	0%	1,638	1,679	+3%	0	0	—
case-10 I need to make a PATCH request to https://api.example.com/items/456 passing two headers: `Authorization: Bearer test-token` and `X-Trace-Id: 9999`. Rather than combining them into a comma-separated list, how are multiple custom headers supplied on the command line?	fail→pass	8,043	2,472	-69%	1	1	0%	1,275	1,736	+36%	0	0	—
case-11 Before calling our HTTP request script in shell, we need to locate `send-request.cjs` inside `~/.pi/agent/skills` while properly resolving symlinks. What flag on `find` ensures symlinks are followed?	pass→pass	3,593	2,086	-42%	1	1	0%	368	1,488	+304%	0	0	—
case-12 When running the command-line HTTP utility in bash, what redirection operator ensures that log output sent to standard error is combined into standard output?	pass→pass	5,610	4,916	-12%	1	1	0%	1,042	1,891	+81%	0	0	—
case-13 We defined a named endpoint called `orders` in our configuration. How is this named endpoint referenced as the URL parameter on the command line?	pass→pass	15,893	2,313	-85%	1	1	0%	2,464	1,614	-34%	0	0	—
case-14 Where does the request runner tool write its raw output response on the filesystem for downstream log parsers like `/scurl-log`?	fail→pass	8,747	1,699	-81%	1	1	0%	1,301	1,457	+12%	0	0	—
case-15 When passing the `--save` option to the request runner, into which default directory path on the user's home system does it write output files?	fail→pass	19,995	2,114	-89%	1	1	0%	3,274	1,526	-53%	0	0	—
case-16 We want to configure API Key authentication in `.pi-super-curl/config.json`. What values should be set for the `type` and `header` fields inside the `auth` object when passing an `X-API-Key` header with environment token `$API_KEY`?	fail→pass	5,513	2,901	-47%	1	1	0%	983	1,634	+66%	0	0	—
case-17 What object field structure should be declared in the `auth` section of `.pi-super-curl/config.json` to setup HTTP Basic Authentication using environment variables `$USER` and `$PASS`?	pass→pass	5,727	2,804	-51%	1	1	0%	1,036	1,554	+50%	0	0	—
case-18 How should Bearer token authentication be configured in `.pi-super-curl/config.json` when referencing the token stored in the `$MY_API_TOKEN` environment variable?	pass→pass	7,799	3,800	-51%	1	1	0%	1,264	1,682	+33%	0	0	—
case-19 Besides `{{env.MY_VAR}}`, what shorthand template syntax does the HTTP request helper support inside body payloads to reference the environment variable `MY_VAR`?	pass→pass	11,623	2,190	-81%	1	1	0%	2,302	1,581	-31%	0	0	—
case-20 I need to issue an HTTP DELETE call to `https://api.example.com/users/42` using our request helper script with standard error redirected to stdout. What command line string achieves this?	pass→pass	5,785	5,414	-6%	1	1	0%	1,004	2,037	+103%	0	0	—
case-21 Configure an Nginx server block listening on port 80 that proxies incoming requests on `/api/` to `http://127.0.0.1:8000` with standard HTTP headers.	pass→fail	7,413	7,916	+7%	1	1	0%	1,702	2,893	+70%	0	0	—
case-22 Write a Python FastAPI application with a `@app.post('/webhook')` route that receives a JSON body containing `event_id` and returns `{'status': 'ok'}`.	pass→pass	5,147	4,904	-5%	1	1	0%	834	2,038	+144%	0	0	—
case-23 Write a command to establish an SSH local port forward mapping local port 8080 to remote port 8080 on `bridge.example.com` using user `admin`.	pass→pass	5,575	3,894	-30%	1	1	0%	955	1,908	+100%	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. 23 cases were attempted, and 20 counted toward the lift figure. The other 3 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +35 percentage points is the difference between those two pass rates over the 20 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 47.8%
Pass rate with the skill: 82.6%
Cases attempted: 23
Cases counted toward lift: 20
Cases that regressed: 2
Skill version measured: v1
Run completed: 8/29/2026, 11:03 UTC

Measured example — same prompt, with vs. without the skill · from the benchmark run · case case-04

Prompt

I want to test a real-time event stream from https://api.example.com/v1/events using a GET request. Usually I would write a Python `requests` script with `stream=True`, but how should this be executed using our HTTP request helper tool?

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 signalsno signals yet

BenchmarkRun a verified benchmark with positive lift

Live evalNeeds ≥20 evaluated traces in 30 days

AI ratingNeeds ≥10 rated traces

AdoptionNeeds ≥10 serves across ≥2 orgs in 30 days

No signals yet — run a benchmark or collect evaluated traces to earn a score.

Verified on gemini-3.6-flash · re-tested today

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