foldrun 20.9.3 → 20.9.5

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/docs/providers.md CHANGED
@@ -226,11 +226,27 @@ The three failures were following the task, not the language:
226
226
  - `default` added an introductory line and ran one list item past its
227
227
  ten-word limit.
228
228
 
229
- So `default` is a sound default for Persian, and an agent whose output must
230
- follow a strict shape (a fixed format, an exact length) earns `max`, or a
231
- `verify:` that checks the shape. One run per case is a small sample — run
232
- the comparison on your own agent's cases with `repeat: 3` before moving a
233
- step to `fast`. The whole comparison cost $0.42 (runs and judges together).
229
+ The whole comparison cost $0.42 (runs and judges together).
230
+
231
+ **The prompt fixed them, not the tier.** The same agent was then told, in
232
+ its prose, the four things it had got wrong: the reply is the deliverable
233
+ (no line about itself, no closing offer); never invent a date, name or
234
+ number; count before answering when the task sets a limit; check that
235
+ subject and verb agree. Run again with `repeat: 2`, `fast` passed 16 / 16
236
+ and `max` 16 / 16, at $0.010 and $0.020 a run ($0.49 for the comparison).
237
+
238
+ A flow around the first agent instead — a `verify: judge:` that the reply
239
+ holds no extra line and meets its limits, with `retry: 2` — did worse:
240
+ `fast` 14 / 16 and `default` 16 / 16, three times slower (about 34 s a
241
+ run). Both of `fast`'s misses were one technical translation that opened
242
+ with a stray English "The", which the verify claim did not ask about. A
243
+ judge checks what its claim names, so name the failure you have seen.
244
+
245
+ So for Persian: write the rules your eval catches into the agent's prose
246
+ first, and keep `fast` if it passes them. A strict shape (a fixed format,
247
+ an exact length) is checked best by a deterministic `verify:` on the
248
+ artefact. Two runs per case is still a small sample: run the comparison on
249
+ your own agent's cases with `repeat: 3` before moving a step to `fast`.
234
250
 
235
251
  ## Why a turn stopped
236
252
 
package/docs/secrets.md CHANGED
@@ -221,7 +221,7 @@ PUT /api/secrets { "name": "…", "value": "…", "sites": ["api.dataforseo.
221
221
 
222
222
  A **remote MCP server**'s `headers:` and `url:` work the same way: bind the
223
223
  key to the server's host and the sandbox's copy of them holds `${NAME}`,
224
- filled by the proxy on the way to that host alone ([MCP](tools#mcp-someone-elses-tools)).
224
+ filled by the proxy on the way to that host alone ([MCP](tools#mcp--someone-elses-tools)).
225
225
 
226
226
  What a program in the sandbox writes is unchanged; it sends the placeholder
227
227
  where the value would go:
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "foldrun",
3
- "version": "20.9.3",
3
+ "version": "20.9.5",
4
4
  "description": "Agents are just folders. Write them, check them, run them — on your machine.",
5
5
  "license": "FSL-1.1-ALv2",
6
6
  "type": "module",
@@ -11,7 +11,7 @@
11
11
  "node": ">=22"
12
12
  },
13
13
  "dependencies": {
14
- "@foldrun/core": "^18.3.2"
14
+ "@foldrun/core": "^18.4.0"
15
15
  },
16
16
  "files": [
17
17
  "bin",