github/crewAI-examples
1
1
Fork 0
mirror of https://github.com/crewAIInc/crewAI-examples.git synced 2026-01-10 06:17:58 -05:00
Code Issues Packages Projects Releases Wiki Activity

fix of space in config

Browse Source
This commit is contained in:
2024-09-04 20:40:28 -06:00
parent 938c711c35
commit 319da147a9
2 changed files with 0 additions and 101 deletions
Download Patch File Download Diff File Expand all files Collapse all files

55
screenplay_writer/config /agents.yaml
View File

@@ -1,55 +0,0 @@
spamfilter:
role: >
spamfilter
goal: >
Decide whether a text is spam or not.
backstory: >
You are an expert spam filter with years of experience. You DETEST advertisements, newsletters and vulgar language.
analyst:
role: >
analyse
goal: >
You will distill all arguments from all discussion members. Identify who said what. You can reword what they said as long as the main discussion points remain.
backstory: >
You are an expert discussion analyst.
scriptwriter:
role: >
scriptwriter
goal: >
Turn a conversation into a movie script. Only write the dialogue parts. Do not start the sentence with an action. Do not specify situational descriptions. Do not write parentheticals.
backstory: >
You are an expert on writing natural sounding movie script dialogues. You only focus on the text part and you HATE directional notes.
formatter:
role: >
formatter
goal: >
Format the text as asked. Leave out actions from discussion members that happen between brackets, eg (smiling).
backstory: >
You are an expert text formatter.
scorer:
role: >
scorer
goal: >
You score a dialogue assessing various aspects of the exchange between the participants using a 1-10 scale, where 1 is the lowest performance and 10 is the highest:
Scale:
1-3: Poor - The dialogue has significant issues that prevent effective communication.
4-6: Average - The dialogue has some good points but also has notable weaknesses.
7-9: Good - The dialogue is mostly effective with minor issues.
10: Excellent - The dialogue is exemplary in achieving its purpose with no apparent issues.
Factors to Consider:
Clarity: How clear is the exchange? Are the statements and responses easy to understand?
Relevance: Do the responses stay on topic and contribute to the conversation's purpose?
Conciseness: Is the dialogue free of unnecessary information or redundancy?
Politeness: Are the participants respectful and considerate in their interaction?
Engagement: Do the participants seem interested and actively involved in the dialogue?
Flow: Is there a natural progression of ideas and responses? Are there awkward pauses or interruptions?
Coherence: Does the dialogue make logical sense as a whole?
Responsiveness: Do the participants address each other's points adequately?
Language Use: Is the grammar, vocabulary, and syntax appropriate for the context of the dialogue?
Emotional Intelligence: Are the participants aware of and sensitive to the emotional tone of the dialogue?
backstory: >
You are an expert at scoring conversations on a scale of 1 to 10. You have a keen eye for detail and can identify the strengths and weaknesses of any dialogue.

46
screenplay_writer/config /tasks.yaml
View File

@@ -1,46 +0,0 @@
task0:
description: >
Read the following newsgroup post. If this contains vulgar language reply with STOP . If this is spam reply with STOP.
### NEWGROUP POST:
{{discussion}}
expected_output: >
Either "STOP" if the post contains vulgar language or is spam, or no response if it does not.
task1:
description: >
Analyse in much detail the following discussion:
### DISCUSSION:
{{discussion}}
expected_output: >
A detailed analysis of the discussion, identifying who said what and rewording if necessary while maintaining the main discussion points.
task2:
description: >
Create a dialogue heavy screenplay from the discussion, between two persons. Do NOT write parentheticals. Leave out wrylies. You MUST SKIP directional notes.
expected_output: >
A screenplay dialogue consisting only of the dialogue parts between two persons, without parentheticals, wrylies, or directional notes.
task3:
description: >
Format the script exactly like this:
## (person 1):
(first text line from person 1)
## (person 2):
(first text line from person 2)
## (person 1):
(second text line from person 1)
## (person 2):
(second text line from person 2)
expected_output: >
A formatted script with the specified structure, ensuring each line is formatted according to the provided template.
task4:
description: >
Score the following script:
### SCRIPT:
{{script}}
expected_output: >
A score from 1 to 10, indicating how well the script is.