@elisym/cli 0.8.1 → 0.9.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +9 -0
- package/dist/index.js +458 -76
- package/dist/index.js.map +1 -1
- package/package.json +2 -2
- package/skills-examples/README.md +17 -15
- package/skills-examples/cheap-summarizer/SKILL.md +26 -0
- package/skills-examples/static-now/SKILL.md +2 -1
- package/skills-examples/static-welcome/SKILL.md +2 -1
- package/skills-examples/uppercase-proxy/SKILL.md +2 -1
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "@elisym/cli",
|
|
3
|
-
"version": "0.
|
|
3
|
+
"version": "0.9.0",
|
|
4
4
|
"description": "CLI agent runner for elisym - provider mode, skills, crash recovery",
|
|
5
5
|
"keywords": [
|
|
6
6
|
"ai-agents",
|
|
@@ -44,7 +44,7 @@
|
|
|
44
44
|
"prepublishOnly": "bun run build && node scripts/preflight-publish.mjs"
|
|
45
45
|
},
|
|
46
46
|
"dependencies": {
|
|
47
|
-
"@elisym/sdk": "~0.
|
|
47
|
+
"@elisym/sdk": "~0.13.0",
|
|
48
48
|
"@solana-program/memo": "~0.11.0",
|
|
49
49
|
"@solana-program/system": "~0.12.0",
|
|
50
50
|
"@solana-program/token": "~0.5.0",
|
|
@@ -6,21 +6,21 @@ These are **provider runtime skills** in elisym's own format (`capabilities`, `p
|
|
|
6
6
|
|
|
7
7
|
## Available skills
|
|
8
8
|
|
|
9
|
-
| Skill | Price
|
|
10
|
-
| ----------------------------------------- |
|
|
11
|
-
| [general-assistant](./general-assistant/) | free
|
|
12
|
-
| [usdc-summarize](./usdc-summarize/) | 0.05 USDC
|
|
13
|
-
| [site-status](./site-status/) | 0.01 USDC
|
|
14
|
-
| [whois-lookup](./whois-lookup/) | 0.01 USDC
|
|
15
|
-
| [github-repo](./github-repo/) | 0.01 USDC
|
|
16
|
-
| [stock-price](./stock-price/) | 0.01 USDC
|
|
17
|
-
| [trending](./trending/) | 0.02 USDC
|
|
18
|
-
| [youtube-summary](./youtube-summary/) | 0.10 USDC
|
|
19
|
-
| [static-welcome](./static-welcome/) | 0.
|
|
20
|
-
| [static-now](./static-now/) | 0.
|
|
21
|
-
| [uppercase-proxy](./uppercase-proxy/) | 0.
|
|
22
|
-
|
|
23
|
-
|
|
9
|
+
| Skill | Price | Mode | What it does |
|
|
10
|
+
| ----------------------------------------- | --------- | -------------- | ----------------------------------------------------------------------------------------- |
|
|
11
|
+
| [general-assistant](./general-assistant/) | free | llm | Summarize, translate, review code, generate text - short answers |
|
|
12
|
+
| [usdc-summarize](./usdc-summarize/) | 0.05 USDC | llm | 2-3 sentence summary of long text |
|
|
13
|
+
| [site-status](./site-status/) | 0.01 USDC | llm + python | HTTP status, response time, SSL validity, redirect chain |
|
|
14
|
+
| [whois-lookup](./whois-lookup/) | 0.01 USDC | llm + python | Domain registrar, dates, name servers, status |
|
|
15
|
+
| [github-repo](./github-repo/) | 0.01 USDC | llm + python | Stars, forks, language, license, last activity for `owner/repo` |
|
|
16
|
+
| [stock-price](./stock-price/) | 0.01 USDC | llm + python | Quote, daily change, volume, 52-week range for a ticker |
|
|
17
|
+
| [trending](./trending/) | 0.02 USDC | llm + python | Top GitHub repos or Reddit posts, ranked |
|
|
18
|
+
| [youtube-summary](./youtube-summary/) | 0.10 USDC | llm + python | Overview, key points, takeaways from a YouTube link (multi-round tools) |
|
|
19
|
+
| [static-welcome](./static-welcome/) | 0.05 USDC | static-file | Sells a fixed Markdown welcome doc. No input box on the buyer side, just a Buy button |
|
|
20
|
+
| [static-now](./static-now/) | 0.01 USDC | static-script | Returns the current UTC timestamp from a 1-line shell script. No LLM, no input |
|
|
21
|
+
| [uppercase-proxy](./uppercase-proxy/) | 0.01 USDC | dynamic-script | Pipes the buyer's text to a script (`tr a-z A-Z`). Skeleton for crypto-paid model proxies |
|
|
22
|
+
|
|
23
|
+
Every paid example is **priced in USDC on Solana devnet**; `general-assistant` is the only free one. Paid skills publish a payment requirement with their capability card and only run after the customer's on-chain transfer is confirmed. To make one free, drop `price` and `token` from its frontmatter; to switch to SOL, set `token: sol` and price in SOL.
|
|
24
24
|
|
|
25
25
|
### Non-LLM modes
|
|
26
26
|
|
|
@@ -77,5 +77,7 @@ Create a folder with a `SKILL.md`. Frontmatter fields:
|
|
|
77
77
|
- `script_timeout_ms` (optional, script modes) - default 60000
|
|
78
78
|
- `tools` (optional, `llm` mode only) - external scripts the LLM can call via `child_process.spawn`
|
|
79
79
|
- `max_tool_rounds` (optional, `llm` mode only) - default 10
|
|
80
|
+
- `provider` + `model` (optional, `llm` mode only) - per-skill LLM override. Must be set together; falls back to agent-level `llm` when omitted. See [`cheap-summarizer/`](./cheap-summarizer/SKILL.md).
|
|
81
|
+
- `max_tokens` (optional, `llm` mode only) - per-skill max-tokens override; independent of `provider`/`model`.
|
|
80
82
|
|
|
81
83
|
Body text after the frontmatter becomes the LLM system prompt. See [`packages/cli/GUIDE.md`](../GUIDE.md) for a full walkthrough and the `youtube-summary` skill for a multi-round tool-use example.
|
|
@@ -0,0 +1,26 @@
|
|
|
1
|
+
---
|
|
2
|
+
name: cheap-summarizer
|
|
3
|
+
description: Summarize text on a smaller, cheaper model than the agent default (0.01 USDC per job, devnet)
|
|
4
|
+
capabilities:
|
|
5
|
+
- summarization
|
|
6
|
+
price: 0.01
|
|
7
|
+
token: usdc
|
|
8
|
+
provider: openai
|
|
9
|
+
model: gpt-5-mini
|
|
10
|
+
max_tokens: 1024
|
|
11
|
+
---
|
|
12
|
+
|
|
13
|
+
You are a concise summarizer. Given input text, return a 2-3 sentence summary in plain text.
|
|
14
|
+
|
|
15
|
+
Rules:
|
|
16
|
+
|
|
17
|
+
- Plain text only, no markdown formatting
|
|
18
|
+
- Match the input language
|
|
19
|
+
- Preserve key facts (names, numbers, dates)
|
|
20
|
+
- Never refuse or ask for clarification - summarize what you have
|
|
21
|
+
|
|
22
|
+
This skill demonstrates per-skill LLM overrides. The agent default model (declared in `elisym.yaml`) is used by every other skill on this agent; only this one routes to `openai / gpt-5-mini` with a tighter `max_tokens` budget. Override fields:
|
|
23
|
+
|
|
24
|
+
- `provider` and `model` must be declared together (or neither). Inheriting just one half from the agent default would produce nonsensical pairs.
|
|
25
|
+
- `max_tokens` overrides independently of the provider/model pair.
|
|
26
|
+
- API keys for any non-default provider come from `secrets.<provider>_api_key` (preferred) or the matching `ANTHROPIC_API_KEY` / `OPENAI_API_KEY` env var.
|