Add structured workflow for managing job applications with /new-application and /populate-application commands. Users can now organize applications in dated folders, store job documents, and have AI analyze requirements to populate strategy documents. Includes bug fix for src/ path references and safe update documentation. OpenSpec: add-application-management 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <noreply@anthropic.com>
148 lines
6.0 KiB
Markdown
148 lines
6.0 KiB
Markdown
# Application Management
|
|
|
|
## ADDED Requirements
|
|
|
|
### Requirement: Application Initialization Command
|
|
|
|
The system SHALL provide a `/new-application` slash command that creates a new application workspace with proper folder structure.
|
|
|
|
#### Scenario: Create new application with minimal info
|
|
|
|
- **WHEN** user runs `/new-application "TechCorp - Senior Developer"`
|
|
- **THEN** system creates folder `src/applications/pending/2025-11-02-TechCorp-Senior-Developer/` with `application.md` template and empty `input/` folder
|
|
|
|
#### Scenario: Create application with date prefix
|
|
|
|
- **WHEN** application folder is created
|
|
- **THEN** folder name is prefixed with current date in `YYYY-MM-DD` format for chronological sorting
|
|
|
|
#### Scenario: Handle special characters in folder names
|
|
|
|
- **WHEN** organization name or job title contains special characters (/, \, :, etc.)
|
|
- **THEN** system sanitizes the folder name by replacing invalid characters with hyphens or removing them
|
|
|
|
#### Scenario: Application template is created
|
|
|
|
- **WHEN** new application is initialized
|
|
- **THEN** `application.md` file is created with template sections (Organization, Job Title, Research, Match Strategy, Key Messages, Tone, etc.)
|
|
|
|
#### Scenario: Input folder is created automatically
|
|
|
|
- **WHEN** new application is initialized
|
|
- **THEN** `input/` subfolder is created for storing job-related documents
|
|
|
|
### Requirement: Application Structure
|
|
|
|
Each application workspace SHALL contain a standardized folder structure and files.
|
|
|
|
#### Scenario: Folder hierarchy is consistent
|
|
|
|
- **WHEN** application is created
|
|
- **THEN** structure is: `src/applications/pending/[date-org-title]/application.md` and `src/applications/pending/[date-org-title]/input/`
|
|
|
|
#### Scenario: Application template has required sections
|
|
|
|
- **WHEN** `application.md` is created
|
|
- **THEN** it includes sections for: Organization Name, Job Title, Job Description Summary, Research Notes, Match Strategy, Key Messages, Tone of Voice, and Document Checklist
|
|
|
|
#### Scenario: Template provides guidance
|
|
|
|
- **WHEN** user opens newly created `application.md`
|
|
- **THEN** each section includes helpful comments explaining what information to include
|
|
|
|
### Requirement: Document Population Command
|
|
|
|
The system SHALL provide a `/populate-application` slash command that analyzes input documents and populates `application.md`.
|
|
|
|
#### Scenario: Populate from current directory
|
|
|
|
- **WHEN** user runs `/populate-application` from within an application folder
|
|
- **THEN** system reads all files in `input/` subfolder and extracts relevant information
|
|
|
|
#### Scenario: Analyze job advertisement
|
|
|
|
- **WHEN** `input/` contains a job advertisement (PDF, DOCX, TXT, MD)
|
|
- **THEN** system extracts: company name, job title, requirements, responsibilities, keywords, and company culture indicators
|
|
|
|
#### Scenario: Analyze email communications
|
|
|
|
- **WHEN** `input/` contains email files or text
|
|
- **THEN** system extracts: recruiter name, company context, timeline, special instructions, and tone expectations
|
|
|
|
#### Scenario: Update application.md with findings
|
|
|
|
- **WHEN** input analysis is complete
|
|
- **THEN** system updates relevant sections of `application.md` with extracted information while preserving any existing manual entries
|
|
|
|
#### Scenario: Ask user for missing information
|
|
|
|
- **WHEN** critical information cannot be extracted from input files (e.g., no clear job title)
|
|
- **THEN** system prompts user to provide missing details before completing population
|
|
|
|
#### Scenario: Handle empty input folder
|
|
|
|
- **WHEN** `/populate-application` is run and `input/` folder is empty or doesn't exist
|
|
- **THEN** system informs user to add job-related documents first and explains what types of files are helpful
|
|
|
|
### Requirement: Profile Integration
|
|
|
|
The application management system SHALL integrate with the existing profile system.
|
|
|
|
#### Scenario: Reference profile during population
|
|
|
|
- **WHEN** `/populate-application` runs
|
|
- **THEN** system reads `src/profile.md` to understand applicant's background for match strategy analysis
|
|
|
|
#### Scenario: Profile validation is preserved
|
|
|
|
- **WHEN** working with applications
|
|
- **THEN** the existing `/validate-profile` requirement remains enforced before document generation (to be implemented in future proposal)
|
|
|
|
#### Scenario: Match applicant to job requirements
|
|
|
|
- **WHEN** populating application
|
|
- **THEN** system identifies which experiences, skills, and projects from profile.md best match the job requirements
|
|
|
|
### Requirement: Claude Code Workflow Instructions
|
|
|
|
The `src/CLAUDE.md` file SHALL include instructions for the application management workflow.
|
|
|
|
#### Scenario: Instructions explain application lifecycle
|
|
|
|
- **WHEN** Claude Code is helping with applications
|
|
- **THEN** it understands the workflow: initialize → add input files → populate → [future: generate documents]
|
|
|
|
#### Scenario: Commands are documented
|
|
|
|
- **WHEN** user asks how to create an application
|
|
- **THEN** Claude Code can explain `/new-application` and `/populate-application` commands
|
|
|
|
#### Scenario: Input file guidance is provided
|
|
|
|
- **WHEN** user asks what files to add to input folder
|
|
- **THEN** Claude Code suggests: job postings (PDF/DOCX/text), recruiter emails, company research notes, or any relevant context
|
|
|
|
### Requirement: File Format Support
|
|
|
|
The population command SHALL support multiple input file formats.
|
|
|
|
#### Scenario: Read PDF files
|
|
|
|
- **WHEN** `input/` contains PDF files
|
|
- **THEN** system extracts text content for analysis (Note: Claude Code has built-in PDF reading capability)
|
|
|
|
#### Scenario: Read text-based formats
|
|
|
|
- **WHEN** `input/` contains TXT, MD, or copied text files
|
|
- **THEN** system reads and analyzes content directly
|
|
|
|
#### Scenario: Read email formats
|
|
|
|
- **WHEN** `input/` contains .eml or plain text email content
|
|
- **THEN** system parses sender, subject, body and extracts relevant application context
|
|
|
|
#### Scenario: Handle unsupported formats gracefully
|
|
|
|
- **WHEN** `input/` contains unsupported file types (e.g., images, videos)
|
|
- **THEN** system skips these files with a warning and processes supported formats
|