token-harness 0.1.23 → 0.1.24
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 +27 -0
- package/package.json +1 -1
- package/sbom.json +3 -3
- package/token-harness.mjs +1173 -54
package/README.md
CHANGED
|
@@ -237,6 +237,33 @@ The app can also preview enabling that guidance in supported user-level Agent Sk
|
|
|
237
237
|
Existing custom skill directories are not silently overwritten or adopted. See
|
|
238
238
|
[RFC 0023](docs/rfcs/0023-guided-agent-skill-install.md).
|
|
239
239
|
|
|
240
|
+
Automatic prompt routing is a separate, opt-in feature in Overview under each detected coding
|
|
241
|
+
agent. When enabled, a native `UserPromptSubmit` hook supplies a short routing policy on every
|
|
242
|
+
prompt; you do not invoke the skill or prefix prompts with a Token Harness command. The root model
|
|
243
|
+
does not change. The agent may delegate one eligible bounded subtask to a native cheaper model,
|
|
244
|
+
then integrate and review its work. Codex asks for `gpt-6-luna`; Claude Code asks for its current
|
|
245
|
+
Haiku alias when available. Claude takes effect in a new session; Codex requires reviewing and
|
|
246
|
+
trusting the hook in `/hooks`. The dashboard separates configured hooks from callbacks observed at
|
|
247
|
+
runtime, and shows model-routing savings only after a paired quality-gated benchmark. Local token
|
|
248
|
+
counts and 5-hour/weekly allowance percentages remain separate measurements. See
|
|
249
|
+
[RFC 0030](docs/rfcs/0030-automatic-native-prompt-routing.md).
|
|
250
|
+
|
|
251
|
+
To plan only the Codex skill from the CLI, exclude provider integrations explicitly, review the
|
|
252
|
+
target path and actions, then apply the returned plan id:
|
|
253
|
+
|
|
254
|
+
```sh
|
|
255
|
+
token-harness plan --harness codex --provider none --agent-skill --json
|
|
256
|
+
token-harness apply --plan <plan-id> --yes
|
|
257
|
+
```
|
|
258
|
+
|
|
259
|
+
To review the native hook directly, use `--agent-routing` to enable it or
|
|
260
|
+
`--disable-agent-routing` to remove the exact entries Token Harness owns:
|
|
261
|
+
|
|
262
|
+
```sh
|
|
263
|
+
token-harness plan --harness claude --provider none --agent-routing --json
|
|
264
|
+
token-harness apply --plan <plan-id> --yes
|
|
265
|
+
```
|
|
266
|
+
|
|
240
267
|
## Advanced CLI
|
|
241
268
|
|
|
242
269
|
Most people do not need these commands. They remain available for automation, debugging and the
|
package/package.json
CHANGED
package/sbom.json
CHANGED
|
@@ -1,14 +1,14 @@
|
|
|
1
1
|
{
|
|
2
2
|
"bomFormat": "CycloneDX",
|
|
3
3
|
"specVersion": "1.5",
|
|
4
|
-
"serialNumber": "urn:uuid:
|
|
4
|
+
"serialNumber": "urn:uuid:dae606f9-46c6-3383-876b-2c09bda8c296",
|
|
5
5
|
"version": 1,
|
|
6
6
|
"metadata": {
|
|
7
7
|
"component": {
|
|
8
8
|
"type": "application",
|
|
9
9
|
"bom-ref": "token-harness",
|
|
10
10
|
"name": "token-harness",
|
|
11
|
-
"version": "0.1.
|
|
11
|
+
"version": "0.1.24",
|
|
12
12
|
"description": "Quota-aware efficiency layer for Claude Code and Codex subscription limits.",
|
|
13
13
|
"licenses": [
|
|
14
14
|
{
|
|
@@ -20,7 +20,7 @@
|
|
|
20
20
|
"hashes": [
|
|
21
21
|
{
|
|
22
22
|
"alg": "SHA-256",
|
|
23
|
-
"content": "
|
|
23
|
+
"content": "dae606f946c63383876b2c09bda8c296da1816849916f5d1774d2f1d84384c47"
|
|
24
24
|
}
|
|
25
25
|
]
|
|
26
26
|
},
|