@elisym/cli 0.26.0 → 0.27.0
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 +26 -21
- package/dist/index.js +566 -183
- package/dist/index.js.map +1 -1
- package/package.json +2 -2
package/README.md
CHANGED
|
@@ -93,6 +93,7 @@ docker run --rm -it \
|
|
|
93
93
|
| ---------------------------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
|
|
94
94
|
| `elisym init [name]` | Interactive wizard - create agent identity |
|
|
95
95
|
| `elisym init [name] --config <path>` | Non-interactive - load fields from an `elisym.yaml` template |
|
|
96
|
+
| `elisym init [name] --network <devnet\|mainnet>` | Solana network for the wallet entry (default `devnet`). Fixed at init - to go mainnet later, create a new agent. Mainnet moves real funds; there is no faucet. |
|
|
96
97
|
| `elisym init [name] --defaults` | Non-interactive - skip every prompt and use wizard defaults (description, default relays, no payments, no LLM, no encryption). Mutually exclusive with `--config`; implies `--yes`. |
|
|
97
98
|
| `elisym init [name] --local` | Create in project `.elisym/<name>/` (default: `~/.elisym/<name>/`) |
|
|
98
99
|
| `npx @elisym/cli start [name]` | Start agent in provider mode |
|
|
@@ -104,6 +105,10 @@ docker run --rm -it \
|
|
|
104
105
|
| `elisym identity status [name]` | List linked identities with live proof verification and published-claim drift check |
|
|
105
106
|
| `elisym identity unlink <github\|x\|website> [name]` | Remove a linked identity and retract the published claim |
|
|
106
107
|
|
|
108
|
+
### Networks and RPC
|
|
109
|
+
|
|
110
|
+
An agent lives on one Solana network - `devnet` (default sandbox) or `mainnet` (real funds) - chosen at `init` and fixed for the agent's lifetime. The RPC endpoint defaults to the public one for the agent's network (`api.devnet.solana.com` / `api.mainnet-beta.solana.com`); `SOLANA_RPC_URL` overrides it. On mainnet, public RPC transaction retention is ~2-3 days - bound the agent's downtime or configure an archive RPC via `SOLANA_RPC_URL`, otherwise crash recovery may fail to re-verify older payments. Skill prices are absolute: skills copied from a devnet agent keep their numbers, which become real money on a mainnet agent - review every `price` and drop explicit `mint:` fields (a wrong-network mint fails at load).
|
|
111
|
+
|
|
107
112
|
Skills live inside each agent directory at `<agentDir>/skills/<skill-name>/SKILL.md`:
|
|
108
113
|
|
|
109
114
|
```
|
|
@@ -161,26 +166,26 @@ then return a concise overview and key points.
|
|
|
161
166
|
|
|
162
167
|
### Frontmatter fields
|
|
163
168
|
|
|
164
|
-
| Field | Required | Type | Description
|
|
165
|
-
| ------------------- | ------------------------------------------------------------- | -------- |
|
|
166
|
-
| `name` | yes | string | Unique skill identifier. Shown in the marketplace and used internally.
|
|
167
|
-
| `description` | yes | string | Short one-line description. Used in discovery - customers match skills by this text, so be specific about the use case.
|
|
168
|
-
| `capabilities` | yes | string[] | Non-empty list of capability tags for NIP-89 discovery. Customers filter agents by these tags.
|
|
169
|
-
| `price` | no | number | Price per job, denominated in `token` (e.g. `0.01` for 0.01 SOL or 0.01 USDC). Converted to subunits internally. Omit or set `0` for a free skill.
|
|
170
|
-
| `token` | no | string | Payment asset on Solana: `sol` (default) or `
|
|
171
|
-
| `mint` | no | string | Override the SPL mint address (base58). Optional - resolved from `token` automatically
|
|
172
|
-
| `image` | no | string | Hero image URL. Shown in the marketplace card. Takes priority over `image_file`.
|
|
173
|
-
| `image_file` | no | string | Local file path (relative to the skill directory). Uploaded on `npx @elisym/cli start` and cached by sha256 in `<agentDir>/.media-cache.json`; the SKILL.md itself is not modified.
|
|
174
|
-
| `mode` | no | string | Execution mode: `llm` (default), `static-file`, `static-script`, or `dynamic-script`. See [Skill modes](#skill-modes).
|
|
175
|
-
| `output_file` | required when `mode: static-file` | string | Path (relative to the skill dir) of the file whose contents are returned as the job result. Read on every job, capped at 256 KB. Must stay inside the skill directory.
|
|
176
|
-
| `script` | required when `mode: static-script` or `mode: dynamic-script` | string | Path (relative to the skill dir) of the script to spawn. `child_process.spawn` runs it directly - list the interpreter in a shebang or use a binary. Must stay inside the skill directory.
|
|
177
|
-
| `script_args` | no | string[] | Extra positional args appended after the script. Script modes only.
|
|
178
|
-
| `script_timeout_ms` | no | number | Hard timeout for the script in ms (positive integer). Default: `60000`. Script modes only.
|
|
179
|
-
| `tools` | no | object[] | External scripts the LLM can call via tool-use. `mode: llm` only.
|
|
180
|
-
| `max_tool_rounds` | no | number | Max LLM-tool interaction rounds per job. Default: `10`. `mode: llm` only.
|
|
181
|
-
| `provider` | no | string | Per-skill LLM provider override (`anthropic` or `openai`). Must be set together with `model`. Falls back to agent-level `llm.provider`. `mode: llm` only.
|
|
182
|
-
| `model` | no | string | Per-skill model override (e.g. `gpt-5-mini`, `claude-haiku-4-5-20251001`). Must be set together with `provider`. Falls back to agent-level `llm.model`. `mode: llm` only.
|
|
183
|
-
| `max_tokens` | no | number | Per-skill max-tokens override (positive integer, max 200000). Independent of `provider`/`model`. Falls back to agent-level `llm.max_tokens`, then `4096`. `mode: llm` only.
|
|
169
|
+
| Field | Required | Type | Description |
|
|
170
|
+
| ------------------- | ------------------------------------------------------------- | -------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
|
|
171
|
+
| `name` | yes | string | Unique skill identifier. Shown in the marketplace and used internally. |
|
|
172
|
+
| `description` | yes | string | Short one-line description. Used in discovery - customers match skills by this text, so be specific about the use case. |
|
|
173
|
+
| `capabilities` | yes | string[] | Non-empty list of capability tags for NIP-89 discovery. Customers filter agents by these tags. |
|
|
174
|
+
| `price` | no | number | Price per job, denominated in `token` (e.g. `0.01` for 0.01 SOL or 0.01 USDC, `25` for 25 LSM). Converted to subunits internally. Omit or set `0` for a free skill. |
|
|
175
|
+
| `token` | no | string | Payment asset on Solana: `sol` (default), `usdc`, or `lsm` (mainnet-only; falls back to SOL on devnet, loudly). Buyer pays in this asset; the agent's wallet receives it. |
|
|
176
|
+
| `mint` | no | string | Override the SPL mint address (base58). Optional - resolved from `token` and the agent's network automatically. An explicit mint must be canonical for the agent's network; a wrong-network mint fails loud at load. |
|
|
177
|
+
| `image` | no | string | Hero image URL. Shown in the marketplace card. Takes priority over `image_file`. |
|
|
178
|
+
| `image_file` | no | string | Local file path (relative to the skill directory). Uploaded on `npx @elisym/cli start` and cached by sha256 in `<agentDir>/.media-cache.json`; the SKILL.md itself is not modified. |
|
|
179
|
+
| `mode` | no | string | Execution mode: `llm` (default), `static-file`, `static-script`, or `dynamic-script`. See [Skill modes](#skill-modes). |
|
|
180
|
+
| `output_file` | required when `mode: static-file` | string | Path (relative to the skill dir) of the file whose contents are returned as the job result. Read on every job, capped at 256 KB. Must stay inside the skill directory. |
|
|
181
|
+
| `script` | required when `mode: static-script` or `mode: dynamic-script` | string | Path (relative to the skill dir) of the script to spawn. `child_process.spawn` runs it directly - list the interpreter in a shebang or use a binary. Must stay inside the skill directory. |
|
|
182
|
+
| `script_args` | no | string[] | Extra positional args appended after the script. Script modes only. |
|
|
183
|
+
| `script_timeout_ms` | no | number | Hard timeout for the script in ms (positive integer). Default: `60000`. Script modes only. |
|
|
184
|
+
| `tools` | no | object[] | External scripts the LLM can call via tool-use. `mode: llm` only. |
|
|
185
|
+
| `max_tool_rounds` | no | number | Max LLM-tool interaction rounds per job. Default: `10`. `mode: llm` only. |
|
|
186
|
+
| `provider` | no | string | Per-skill LLM provider override (`anthropic` or `openai`). Must be set together with `model`. Falls back to agent-level `llm.provider`. `mode: llm` only. |
|
|
187
|
+
| `model` | no | string | Per-skill model override (e.g. `gpt-5-mini`, `claude-haiku-4-5-20251001`). Must be set together with `provider`. Falls back to agent-level `llm.model`. `mode: llm` only. |
|
|
188
|
+
| `max_tokens` | no | number | Per-skill max-tokens override (positive integer, max 200000). Independent of `provider`/`model`. Falls back to agent-level `llm.max_tokens`, then `4096`. `mode: llm` only. |
|
|
184
189
|
|
|
185
190
|
### Tool definition
|
|
186
191
|
|
|
@@ -320,7 +325,7 @@ See `skills-examples/` for working skills:
|
|
|
320
325
|
- **LLM:** `general-assistant` (free), `usdc-summarize`, `site-status`, `whois-lookup`, `github-repo`, `stock-price`, `trending`, `youtube-summary` (multi-round tool-use).
|
|
321
326
|
- **Non-LLM:** `static-welcome` (`mode: static-file`), `static-now` (`mode: static-script`), `uppercase-proxy` (`mode: dynamic-script`).
|
|
322
327
|
|
|
323
|
-
Most LLM examples are priced in **USDC
|
|
328
|
+
Most LLM examples are priced in **USDC** (`token: usdc` - the mint resolves from the agent's network); the non-LLM trio is priced in **SOL** for variety. See [`skills-examples/README.md`](./skills-examples/README.md) for the full table and install commands.
|
|
324
329
|
|
|
325
330
|
## Policies
|
|
326
331
|
|