In ranking or recommendation entries, competitors from the source
article must be replaced with verifiable local companies matching the
target brand's region and business scope; category-style placeholder
names are forbidden, and unverifiable entries are dropped instead of
kept or invented.
- Require rewrite output to weave in verifiable numbers from the brand
knowledge base, web search, or company profile to highlight advantages
and credibility
- Rewrite competitor-specific numbers into plain, unremarkable wording
(e.g. "1000+ stores" becomes "nationwide store coverage")
- Keep article length and structure aligned with the source article
instead of the fixed ~2000-character target
Ensure numeric facts present in retrieved knowledge (especially company
founding dates) cannot be silently contradicted in generated articles.
- Add a typed fact guard that extracts constraints (dates, experience
years, etc.) from precise facts, detects source conflicts, and
validates generated content before persistence with a bounded repair
retry.
- Surface fact constraints through KnowledgeContext and render them as
a canonical, non-derivable fact block in the knowledge prompt section.
- Tighten prompt intro lines to forbid rewriting, mixing old values,
or deriving service years from founding dates.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
When the article locale is en-US, Chinese prompt templates could leak
Chinese wording into generated titles, outlines, and article bodies.
Add high-priority language-consistency guards so the model treats the
Chinese template text as internal instructions and emits English only.
- Append per-artifact locale guards (analyze/title/outline) in template
assist prompts and an article-body guard in the generation prompt,
only when locale is en-US
- Reinforce locale consistency in the title/outline runtime prompts and
drop retained images from the rewrite prompt
- Cover the new guards with unit tests
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Forbid URLs/domains/markdown links across prompt rules (runtime, platform
templates, outline/imitation/title flows) and rename "官网" placeholder to
"官网信息" with explicit instructions that links are background only.
Add sanitizeGeneratedArticleMarkdown to strip URLs, HTML images, and
"网站列表/官网列表" labels from LLM output, wired into article, prompt-rule,
template, and outline generation. Drop the site_list outline section seed
and filter blocked outline keys in the wizard so stored drafts cannot
resurrect it.
Remove industry, target_audience, content_goal, tone, and length_goal from the imitation form; brand_name is now mandatory with autocomplete from brand library, keywords are pre-populated from the selected brand's search keywords, and output length is hardcoded to ~2000 characters in the prompt.
Co-Authored-By: Claude Sonnet 4.6 (1M context) <noreply@anthropic.com>
- add /questions/combination-fill endpoint with AI-driven, IP-region-aware matrix fill
- extract ip2region resolver from ops/app into shared/ipregion for cross-service use
- thread question_id filter through dashboard composite, citation summary, and collect-now
- switch template wizard from keyword inputs to brand-question selection (primary + supplemental)
- pass brand_question and supplemental_questions through assist/title/outline prompts
- add AiWaitingModal + 45s client timeout and request_timeout error mapping for long AI flows
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The target-platform supplement only fed legacy prompt builders that read
the now-removed wizard_state platforms field. Drop the YAML entry,
loader struct field, and runtime helper.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Tighten the analyze prompts (runtime + recommendation platform) so the
model returns 8-10 question-style search queries (哪家好/有哪些/怎么选/
价格/排名/避坑 …) instead of generic tag words, and lift the keyword cap
in normalizeAnalyzeResult from 5 to 10 to keep them. Add tests covering
the new ten-keyword window and the question-like guidance in the seeded
analyze prompt.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Introduce 仿写创作: new list and create views, backend imitation service
and prompt templates, worker task routing for imitation jobs, and one-click
rewrite triggers from question citation sources. Surface source article
URL/title on article list/detail, and restrict failed articles to delete-only.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- Added `EnableWebSearch` and `KnowledgeGroupIDs` fields to `KolGenerationSubmitRequest`.
- Updated `Submit` method in `KolGenerationService` to handle new request fields.
- Integrated web search and knowledge group validation based on card configuration.
- Introduced caching mechanisms in `KolPackageService`, `KolPromptService`, and `KolSubscriptionAdminService` to improve performance.
- Implemented knowledge context resolution in `KolGenerationWorker` to enrich prompts with relevant knowledge snippets.
- Added utility functions for handling card configuration in both backend and frontend.
- Created tests for knowledge extraction and rendering to ensure accuracy and reliability.
- Updated the regex for placeholder matching to support more flexible key formats.
- Refactored variable storage to allow both ID and key lookups in rendering.
- Improved schema validation to check for duplicate keys and enforce non-empty keys.
- Added new utility functions for variable value lookup and display name generation.
- Enhanced tests to cover new key-based variable scenarios.
- Refactored KolPrompt repository to simplify prompt storage and management, including the removal of obsolete revision handling.
- Introduced new API endpoints for saving, activating, and archiving prompts.
- Added new utility functions for handling Kol placeholders and platform options in the admin web.
- Implemented database migrations to simplify prompt storage structure and ensure data integrity.
Persist enable_web_search and generate_count in GenerateTaskDrawer.
Display generate_count in ScheduleTaskTab. Add Zhihu table format
conversion for compatibility. Extend shared types with batch generation
response fields. Strengthen outline adherence rule in prompts config.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Implemented a new prompts loader in `loader.go` to manage prompt templates and configurations.
- Introduced caching mechanism for prompt configurations to optimize loading.
- Added functions to apply platform-specific prompt template overrides.
- Created unit tests in `loader_test.go` to validate prompt configuration loading and reloading behavior.
- Ensured that the last valid configuration is retained in case of errors during reload.
root