61f13a806d
test(interviews): end-to-end pipeline test + content-aware LLM stubs for all 4 subagents
...
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-23 12:40:53 +02:00
52bae0a3da
feat(interviews): Flask blueprint /api/interview with task-based async + CSV export
...
Add /api/interview blueprint with POST pre/post/rerun, GET status/results/synthesis/export.csv endpoints. Background tasks tracked by UUID in module-level dict. Add register_blueprints() helper to api/__init__.py and wire app factory through it. Add UPLOADS_DIR to Config with env-override default.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-23 12:37:04 +02:00
bc07170dbf
feat(interviews): persona + Zep memory adapters bridging existing services to interview subsystem
...
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-23 12:34:03 +02:00
d79c81d2b7
feat(interviews): synthesiser emits cross-method report + tidy CSV + limitations section
...
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-23 12:32:14 +02:00
3322bcb20c
feat(interviews): on_ready / on_completed hook registry on SimulationManager
...
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-23 12:29:30 +02:00
b3e2039817
feat(interviews): orchestrator with two-phase lifecycle, parallel fan-out, isolated failures
...
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-23 12:27:47 +02:00
cca67365b9
feat(interviews): Zep writer adapts add_activity/add_text_episode for per-agent + aggregate episodes
...
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-23 12:24:33 +02:00
998cf1ac27
feat(interviews): JSONL/JSON storage layout with run_id directories and latest pointer
...
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-23 12:23:06 +02:00
ae4941df8e
feat(interviews): scenario subagent with 4 futures × 4 dimensions + polarity matrix
...
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-23 12:21:21 +02:00
5d7111b54e
feat(interviews): Delphi subagent (3 rounds: open, rate, revise) + convergence metrics
...
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-23 12:19:07 +02:00
75762ccc18
feat(interviews): diversity subagent with Q-sort + 6 Likert axes + PCA/k-means typology
...
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-23 12:16:21 +02:00
0fcb815cde
feat(interviews): longitudinal subagent + 12-item Likert instrument
...
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-23 12:12:46 +02:00
289a0cff56
feat(interviews): StakeholderInterviewer base with in-character prompting and schema retry
...
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-23 12:10:01 +02:00
eb3c3629c1
feat(interviews): LLM stub mode for deterministic CI tests
...
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-23 12:08:29 +02:00
29be754ff4
feat(interviews): YAML instrument loader with pydantic validation and hash freezing
...
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-23 12:06:52 +02:00
f1898b4eac
feat(interviews): add pydantic models for instruments and responses
...
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-23 12:04:45 +02:00
071f8b5c4c
feat(interviews): add interview config keys (token budget, workers, language, stub mode)
...
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-23 12:02:33 +02:00
f63bc5542a
chore(interviews): add deps and pytest scaffold for interview subsystem
...
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-23 12:00:09 +02:00
af71244974
Merge pull request #428 from Ghostubborn/feat/i18n
...
feat(i18n): 添加多语言切换功能，支持中英文
2026-04-02 14:27:04 +08:00
f2404903d6
fix(i18n): validate Accept-Language header against registered locales
2026-04-02 14:20:15 +08:00
24e9bee5be
feat(i18n): replace all user-visible Chinese logger messages in zep_tools.py
...
These are shown to users via ConsoleLogger in the report page.
2026-04-01 17:46:39 +08:00
e79569ab4f
feat(i18n): replace all user-visible Chinese in report_agent.py
...
Covers ReportLogger message fields and logger messages shown via ConsoleLogger.
2026-04-01 17:44:52 +08:00
e3350a919d
fix(graph): enforce PascalCase for entity names and SCREAMING_SNAKE_CASE for edge names in ontology validation
2026-04-01 17:42:27 +08:00
380e456d41
fix(i18n): replace hardcoded Chinese stage names in simulation prepare SSE
2026-04-01 17:31:00 +08:00
0e55e4cf6b
feat(i18n): replace remaining Chinese in config generator and profile generator
...
Also update simulation prompts to be locale-neutral for timezone/schedule.
2026-04-01 17:19:12 +08:00
7c07237544
fix(i18n): pass locale to background threads via thread-local storage
...
Background threads (graph building, simulation prep, report generation,
profile generation) now inherit the requesting user's locale preference.
Previously these fell back to 'zh' because Flask request context was
unavailable in spawned threads.
2026-04-01 16:55:51 +08:00
592ee52f59
feat(i18n): replace remaining hardcoded Chinese in progress callbacks
2026-04-01 16:53:29 +08:00
da2490ec31
fix(i18n): protect JSON field values from language instruction in config generator
...
Ensure poster_type stays PascalCase English and stance stays English enum
values regardless of language setting. Only natural language fields follow
the user's language preference.
2026-04-01 16:41:22 +08:00
97aa58384e
fix(i18n): ensure ontology names stay PascalCase regardless of language setting
...
The language instruction was causing LLM to change entity/relation naming
conventions. Now explicitly enforce PascalCase/UPPER_SNAKE_CASE for technical
identifiers while only applying language preference to description fields.
2026-04-01 16:40:17 +08:00
9d43b77511
feat(i18n): replace hardcoded Chinese in backend SSE progress messages
2026-04-01 16:32:10 +08:00
f75c6487b3
fix(i18n): replace remaining hardcoded language directives in LLM prompts
...
- oasis_profile_generator: replace hardcoded "使用中文" with dynamic get_language_instruction()
- ontology_generator: remove hardcoded "（中文）" from schema annotation
- report_agent: replace Chinese-specific language consistency rules with language-neutral ones
- zep_tools: dynamically select quote style based on locale
2026-04-01 15:55:04 +08:00
74f673a238
feat(i18n): replace hardcoded Chinese in backend API responses with t() calls
2026-04-01 15:32:24 +08:00
8f6110df0f
feat(i18n): inject language instruction into LLM system prompts
2026-04-01 15:24:12 +08:00
0c18e1aeca
feat(i18n): add backend translation utility with shared locale files
2026-04-01 15:22:14 +08:00
985f89f49a
fix: resolve 500 error caused by <think> tags and markdown code fences in content field from reasoning models like MiniMax/GLM
2026-03-06 00:30:31 +08:00
da6548e96f
feat(graph): implement pagination for fetching nodes and edges; add utility functions for streamlined data retrieval
2026-02-27 15:53:29 +08:00
25aa4f75d2
fix(report_agent): refine tool call handling and response validation; enforce strict separation between tool calls and final answers
2026-02-24 17:47:44 +08:00
08ec856a58
fix(report_agent): update max_agents parameter description and enforce maximum limit of 10 agents
2026-02-14 18:35:05 +08:00
ddd9ff2479
feat(report_agent): update report language consistency guidelines; ensure all quoted content is translated to the report language for clarity
2026-02-14 18:24:03 +08:00
7601d78fd4
feat(report_agent): enhance interview text processing and response handling; improve quote extraction and formatting for better clarity
2026-02-14 16:56:48 +08:00
dc0a9261d1
feat(report_agent): add detailed tool descriptions and prompts for future prediction report generation
2026-02-14 15:16:17 +08:00
d2041f6fb8
fix(report_agent): update description of insight_forge tool to remove "最强大" and enhance clarity
2026-02-14 14:48:23 +08:00
0a59bace92
fix(report_agent): increase minimum tool call requirement from 2 to 3 per chapter and enhance user prompts to encourage diverse tool usage
2026-02-06 19:37:52 +08:00
e004fe8f14
fix(report_agent): update tool call requirements in content generation to allow up to 5 tool calls per chapter and clarify user prompts for insufficient data
2026-02-06 18:34:19 +08:00
f9abaf8e9f
refactor(report_agent, Step4Report): simplify logging and remove subsection handling; update UI to reflect changes in section content generation
2026-02-06 18:13:30 +08:00
54f1291967
fix(report_agent): handle None responses from LLM during content generation and enforce fallback behavior
2026-01-29 17:08:39 +08:00
40f7035947
fix(config): enable overriding of environment variables when loading .env file
2026-01-23 16:20:24 +08:00
390c120fef
fix(file_parser): handle non-UTF-8 encoded text files with automatic encoding detection
2026-01-22 18:28:37 +08:00
56b8babf17
feat(ZepGraphMemoryUpdater): add platform display name mapping and logging enhancements.
2026-01-16 09:00:10 +08:00
e25d2e38f4
feat(SimulationAPI): add function to retrieve latest report ID for a given simulation
...
- Implemented `_get_report_id_for_simulation` to find the most recent report ID associated with a simulation ID by scanning the reports directory.
- Updated `get_simulation_history` to include the retrieved report ID in the response, enhancing the simulation data returned to the client.
2026-01-09 16:04:35 +08:00
Christian Moellmann
BaiFu
ghostubborn