Remove "generate project" for clarity

services/testing/README.md

@@ -50,7 +50,7 @@ Two rules of thumb:
   the service payload, calls the service via `ApolloClient`, and runs the judge.
   For inspection, three artifacts are written to a `tmp/` folder next to the
   spec file:
-  - `<spec_id>.yaml` — any project YAML in the response (`response_yaml`,
+  - `<spec_id>.yaml` — any workflow YAML in the response (`response_yaml`,
     `workflow_yaml`, `content_yaml`, or a `workflow_yaml` attachment).
   - `<spec_id>.txt` — the response text, prefixed with the agent path
     (e.g. `agents: router -> planner -> job_code_agent`).

services/testing/spec_collector.py

@@ -105,7 +105,7 @@ def runtest(self):
             response, spec.id, self.run_index, spec.runs, tmp_dir, experiment
         )
         if yaml_path is not None:
-            print(f"  ✓ project YAML saved to {yaml_path}")
+            print(f"  ✓ workflow YAML saved to {yaml_path}")
 
         text_path = _capture_response_text(
             response, spec.id, self.run_index, spec.runs, tmp_dir, experiment
@@ -186,7 +186,7 @@ def _build_payload(spec: Spec) -> dict:
 
 
 def _extract_yaml_from_response(response: dict) -> str | None:
-    """Pull the full project YAML out of a service response, if any.
+    """Pull the full workflow YAML out of a service response, if any.
 
     Different services use different keys:
       - workflow_chat: `response_yaml`
@@ -222,7 +222,7 @@ def _capture_response_yaml(
     output_dir: Path,
     experiment: str = "",
 ) -> Path | None:
-    """Write the response's project YAML to `output_dir/<spec_id>.yaml`.
+    """Write the response's workflow YAML to `output_dir/<spec_id>.yaml`.
 
     For multi-run specs, appends `__run-N` to the filename so each run is
     preserved. `experiment` (already a `__<label>` suffix or '') is appended

services/workflow_chat/README.md

@@ -86,7 +86,7 @@ alternatively, the conversation history, given the error message in `errors`. It
 will return an answer whatever information is given, but it is best to give it
 the `existing_yaml` and conversation `history`.
 
-See the `fix_yaml_error_system_prompt` in `gen_project_prompts.yaml` to see the
+See the `fix_yaml_error_system_prompt` in `prompts.yaml` to see the
 exact instructions.
 
 ````json

services/workflow_chat/config_loader.py

@@ -1,7 +1,7 @@
 import yaml
 
 class ConfigLoader:
-    def __init__(self, config_path="gen_project_config.yaml", prompts_path="gen_project_prompts.yaml"):
+    def __init__(self, config_path="config.yaml", prompts_path="prompts.yaml"):
         self.config = self.load_yaml(config_path)
         self.prompts = self.load_yaml(prompts_path)
 

services/workflow_chat/gen_project_prompt.py → services/workflow_chat/prompt.py

@@ -7,8 +7,8 @@
 from .config_loader import ConfigLoader
 
 base_dir = os.path.dirname(os.path.abspath(__file__))
-config_path = os.path.join(base_dir, "gen_project_config.yaml")
-prompts_path = os.path.join(base_dir, "gen_project_prompts.yaml")
+config_path = os.path.join(base_dir, "config.yaml")
+prompts_path = os.path.join(base_dir, "prompts.yaml")
 
 config_loader = ConfigLoader(config_path=config_path, prompts_path=prompts_path)
 config = config_loader.config

services/workflow_chat/gen_project_prompts.yaml → services/workflow_chat/prompts.yaml

@@ -1,4 +1,4 @@
-prompts_version: 1.4.0
+prompts_version: 1.4.1
 prompts:
   main_system_prompt: |
     You are an expert assistant for the OpenFn workflow automation platform.
@@ -11,7 +11,7 @@ prompts:
     2. Determine appropriate adaptors for each job
     3. Set up proper trigger mechanisms (webhook or cron). Note that ONLY ONE node/job can come from the trigger.
     4. Create the connections (edges) between jobs
-    5. Generate a valid project.yaml file that follows OpenFn's structure
+    5. Generate a valid workflow YAML that follows OpenFn's structure
 
     When describing your overall purpose to the user, use simple language and explain your task is to help them create an OpenFn workflow. Avoid mentioning YAMLs, as the user does not see the YAML, but a chart drawn based on it.
 
@@ -29,11 +29,11 @@ prompts:
 
   yaml_structure_with_ids: |
 
-    ## OpenFn Project.yaml Structure
+    ## OpenFn Workflow YAML Structure
 
-    A valid project.yaml must follow this structure:
+    A valid workflow YAML must follow this structure:
     ```
-    name: open-project
+    name: my-workflow
     jobs:
       job-one:
         id: '__ID_JOB_job-one__'
@@ -83,11 +83,11 @@ prompts:
     Webhook triggers may include optional sub-keys (e.g. `custom_path`, `webhook_reply`, `webhook_response_config`) — preserve any such sub-keys EXACTLY, including ones not described here. See the webhook sub-key sections below.
 
   yaml_structure_without_ids: |
-    ## OpenFn Project.yaml Structure
+    ## OpenFn Workflow YAML Structure
 
-    A valid project.yaml must follow this structure:
+    A valid workflow YAML must follow this structure:
     ```
-    name: open-project
+    name: my-workflow
     jobs:
       job-one:
         name: First Job

services/workflow_chat/tests/unit/gen_project/test_prompt_build.py → services/workflow_chat/tests/unit/prompt/test_prompt_build.py

@@ -1,6 +1,6 @@
 import pytest
 from name_rules import UNICODE_FLAG_ENV, describe_rule_for_prompt
-from workflow_chat.gen_project_prompt import build_prompt
+from workflow_chat.prompt import build_prompt
 
 
 def test_build_prompt_normal_mode():
@@ -58,7 +58,7 @@ def test_build_prompt_subagent_strips_inspector_instruction():
 
     # The go-elsewhere phrasing must not be in context at all, in any of the
     # prompt sections it appears in — if this fails, the sentence in
-    # gen_project_prompts.yaml and the replace() in build_prompt have drifted
+    # prompts.yaml and the replace() in build_prompt have drifted
     assert "navigate to the specific job's code page in the Inspector" not in system_msg
     assert "DECLINE" not in system_msg
     assert 'If the user asks for job code, set "handover"' in system_msg
@@ -143,16 +143,16 @@ def test_a_prompt_that_drops_the_name_rule_token_is_rejected_loudly(
     The prompt would then state no naming rule at all while the sanitizer
     carried on enforcing one, and the model would be left guessing.
     """
-    from workflow_chat import gen_project_prompt
+    from workflow_chat import prompt as prompt_module
 
     monkeypatch.setattr(
-        gen_project_prompt.config_loader,
+        prompt_module.config_loader,
         "get_prompt",
         lambda name: "no token here {adaptors}" if name == "general_knowledge" else "x",
     )
 
     with pytest.raises(ValueError, match="did not render the step-name rule"):
-        gen_project_prompt.build_prompt(content="Create a workflow")
+        prompt_module.build_prompt(content="Create a workflow")
 
 def test_build_prompt_describes_webhook_custom_path():
     system_msg, _ = build_prompt(

services/workflow_chat/workflow_chat.py

@@ -8,7 +8,7 @@
 from models import resolve_model
 
 _dir = os.path.dirname(os.path.abspath(__file__))
-with open(os.path.join(_dir, "gen_project_config.yaml")) as _f:
+with open(os.path.join(_dir, "config.yaml")) as _f:
     _service_config = yaml.safe_load(_f)
 
 _MODEL = resolve_model(_service_config.get("model", "claude-sonnet"))
@@ -72,7 +72,7 @@
 )
 from util import ApolloError, create_logger, add_page_prefix, APOLLO_VERSION
 from yaml_utils import _remove_ids
-from .gen_project_prompt import build_prompt
+from .prompt import build_prompt
 from workflow_chat.available_adaptors import get_available_adaptors
 from streaming_util import (
     StreamManager,