cxas-cuj-report-generator

Name: cxas-cuj-report-generator
Author: GoogleCloudPlatform

27installs64forks

How to install cxas-cuj-report-generator?

googlecloudplatform/cxas-scrapi/cxas-cuj-report-generator

$npx -y skills add GoogleCloudPlatform/cxas-scrapi --skill cxas-cuj-report-generator

Installs into the current project.

Summary

Automates the ingestion of customer requirement documents such as diagrams, BRDs, code etc., synthesizes high-fidelity natural transcripts, and compiles them into highly interactive, responsive Critical User Journey (CUJ) reports.

Files · 1

View on GitHub

SKILL.md

1	# Critical User Journey (CUJ) Transcript & Report Generator Skill
2
3	Use this skill when asked to extract dialogue transcripts or compile interactive
4	Critical User Journey (CUJ) reports from a directory of customer requirement
5	documents (such as diagrams, BRDs, code etc.).
6
7	## Core Protocols
8
9	To ensure 100% coverage and zero data loss, you MUST follow these core rules:
10
11	* Robust Extraction: Follow the protocol defined in the
12	`cxas-protocol-robust-extraction` skill.
13	* Two-Phase Ingestion: Follow the protocol defined in the
14	`cxas-protocol-two-phase-ingestion` sub-protocol inside
15	`protocols/cxas-protocol-two-phase-ingestion/`.
16	* Checklist Mandate: The orchestrator and all subagents MUST follow the
17	`agent-protocol-checklist` protocol to maintain a local
18	`task_checklist.json` file, ensuring they track their progress and not lose
19	coverage during execution.
20	* Orchestrator Delivery Assurance: The orchestrator MUST act
21	as a strict, independent Delivery Auditor. BEFORE closing subagents,
22	terminating the watchdog, or reporting campaign success to the user, the
23	orchestrator MUST physically verify the existence, size bounds, and schema
24	compliance of all registered deliverables (specifically
25	`gecx_customer_report.html` and `gecx_cuj_report.html`) on disk. Under no
26	circumstances may the orchestrator assume completion without executing a
27	physical file-presence check.
28	* Auditing: The orchestrator MUST periodically check the subagent's
29	scratch directory to ensure the `task_checklist.json` file is being created
30	and maintained. If the file is missing or not updated, the orchestrator MUST
31	terminate the subagent and respawn it with stronger enforcement
32	instructions.
33
34	## Core Workflow Steps
35
36	Follow this 5-step structured workflow to execute the task:
37
38	1. Scoping & Type Discovery: Prepare the environment and identify required
39	skills.
40
41	* Access Files: Ensure you have access to the source artifacts in your
42	local workspace.
43	* Tip (Drive Links): If the source is a Google Drive link or
44	folder ID, you MUST use the `gdrive` skill to access them.
45	* Detect Inventory Types: To identify framework signatures and map
46	them to correct Ingestors, you MUST use the framework detector agent
47	defined in `agents/framework_detector.md`. Using this agent, scan the
48	input files to inventory all file extensions and detect potential
49	frameworks. Spawn parallel Framework Detector subagents to scan
50	partitions of the file tree.
51	* Map Ingestors: Use the scoping report generated by the Framework
52	Detector to select or create the correct specialized skills in
53	`ingestors/frameworks/` or `ingestors/files/`.
54
55	* Precedence Rule: Framework-specific ingestors take precedence
56	over generic file-extension ingestors (e.g., use
57	`ingestors/frameworks/adk/` instead of `ingestors/files/py/` if both
58	apply).
59
60	2. Discovery: Spawn specialized expert subagents based on the discovered
61	types to identify sub-intents (see the `agents/` directory for role
62	definitions). Dynamically discover and use specialized ingestor skills in
63	`ingestors/frameworks/` and `ingestors/files/`.
64
65	* Mandatory Handoff: Subagents MUST report back:
66
67	1. Frameworks detected,
68	2. File types parsed, and
69	3. Any files/patterns skipped as out-of-scope.
70
71	* Exhaustive Use: Use all relevant ingestors by applying the most
72	specific one applicable to each file.
73
74	* Fallback: If no specialized ingestor exists for an out-of-scope file
75	type, the orchestrator MUST delegate the analysis:
76
77	1. Spawn Analyzer: Spawn a specialized Analysis Subagent to
78	inspect a sample of the unknown file.
79	2. Research: Instruct the subagent to search online or in internal
80	documentation for format standards if the structure is not clear.
81	3. Report & Codify: The subagent must report the best parsing
82	strategy back to the orchestrator and SHOULD attempt to create a new
83	specialized skill in `ingestors/frameworks/` or `ingestors/files/`
84	to capture this knowledge.
85
86	3. Exhaustion: Loop until no new intents are found.
87
88	4. Clustering: Group into Parent CUJs. To ensure consistent and accurate
89	category discovery:
90
91	* Noise Reduction: Do NOT pass full objects with raw transcripts or
92	code.
93	* Summary Format: Provide a clean YAML list with `id`, `name`
94	(stripped of technical tag

People who liked this also viewed

6 picks

Type

cxas-cuj-report-generator

27installs64forks

View on GitHub

How to install cxas-cuj-report-generator?

googlecloudplatform/cxas-scrapi/cxas-cuj-report-generator

$npx -y skills add GoogleCloudPlatform/cxas-scrapi --skill cxas-cuj-report-generator

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Critical User Journey (CUJ) Transcript & Report Generator Skill
2
3	Use this skill when asked to extract dialogue transcripts or compile interactive
4	Critical User Journey (CUJ) reports from a directory of customer requirement
5	documents (such as diagrams, BRDs, code etc.).
6
7	## Core Protocols
8
9	To ensure 100% coverage and zero data loss, you MUST follow these core rules:
10
11	* Robust Extraction: Follow the protocol defined in the
12	`cxas-protocol-robust-extraction` skill.
13	* Two-Phase Ingestion: Follow the protocol defined in the
14	`cxas-protocol-two-phase-ingestion` sub-protocol inside
15	`protocols/cxas-protocol-two-phase-ingestion/`.
16	* Checklist Mandate: The orchestrator and all subagents MUST follow the
17	`agent-protocol-checklist` protocol to maintain a local
18	`task_checklist.json` file, ensuring they track their progress and not lose
19	coverage during execution.
20	* Orchestrator Delivery Assurance: The orchestrator MUST act
21	as a strict, independent Delivery Auditor. BEFORE closing subagents,
22	terminating the watchdog, or reporting campaign success to the user, the
23	orchestrator MUST physically verify the existence, size bounds, and schema
24	compliance of all registered deliverables (specifically
25	`gecx_customer_report.html` and `gecx_cuj_report.html`) on disk. Under no
26	circumstances may the orchestrator assume completion without executing a
27	physical file-presence check.
28	* Auditing: The orchestrator MUST periodically check the subagent's
29	scratch directory to ensure the `task_checklist.json` file is being created
30	and maintained. If the file is missing or not updated, the orchestrator MUST
31	terminate the subagent and respawn it with stronger enforcement
32	instructions.
33
34	## Core Workflow Steps
35
36	Follow this 5-step structured workflow to execute the task:
37
38	1. Scoping & Type Discovery: Prepare the environment and identify required
39	skills.
40
41	* Access Files: Ensure you have access to the source artifacts in your
42	local workspace.
43	* Tip (Drive Links): If the source is a Google Drive link or
44	folder ID, you MUST use the `gdrive` skill to access them.
45	* Detect Inventory Types: To identify framework signatures and map
46	them to correct Ingestors, you MUST use the framework detector agent
47	defined in `agents/framework_detector.md`. Using this agent, scan the
48	input files to inventory all file extensions and detect potential
49	frameworks. Spawn parallel Framework Detector subagents to scan
50	partitions of the file tree.
51	* Map Ingestors: Use the scoping report generated by the Framework
52	Detector to select or create the correct specialized skills in
53	`ingestors/frameworks/` or `ingestors/files/`.
54
55	* Precedence Rule: Framework-specific ingestors take precedence
56	over generic file-extension ingestors (e.g., use
57	`ingestors/frameworks/adk/` instead of `ingestors/files/py/` if both
58	apply).
59
60	2. Discovery: Spawn specialized expert subagents based on the discovered
61	types to identify sub-intents (see the `agents/` directory for role
62	definitions). Dynamically discover and use specialized ingestor skills in
63	`ingestors/frameworks/` and `ingestors/files/`.
64
65	* Mandatory Handoff: Subagents MUST report back:
66
67	1. Frameworks detected,
68	2. File types parsed, and
69	3. Any files/patterns skipped as out-of-scope.
70
71	* Exhaustive Use: Use all relevant ingestors by applying the most
72	specific one applicable to each file.
73
74	* Fallback: If no specialized ingestor exists for an out-of-scope file
75	type, the orchestrator MUST delegate the analysis:
76
77	1. Spawn Analyzer: Spawn a specialized Analysis Subagent to
78	inspect a sample of the unknown file.
79	2. Research: Instruct the subagent to search online or in internal
80	documentation for format standards if the structure is not clear.
81	3. Report & Codify: The subagent must report the best parsing
82	strategy back to the orchestrator and SHOULD attempt to create a new
83	specialized skill in `ingestors/frameworks/` or `ingestors/files/`
84	to capture this knowledge.
85
86	3. Exhaustion: Loop until no new intents are found.
87
88	4. Clustering: Group into Parent CUJs. To ensure consistent and accurate
89	category discovery:
90
91	* Noise Reduction: Do NOT pass full objects with raw transcripts or
92	code.
93	* Summary Format: Provide a clean YAML list with `id`, `name`
94	(stripped of technical tag

People who liked this also viewed

6 picks

Type