菜单
Data-retention and memory-hygiene summarization. Compress one calendar day of source-table rows, or a cluster of older same-source memories, into a single dense paragraph. Use when bounding storage growth while preserving the gist for retrieval.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
TTS pronunciation configuration. Controls how written text is normalized to natural spoken English before being passed to the Speaches/Kokoro TTS engine. Two DB-configurable app_settings keys let operators add or override pronunciation rules without code changes.
Image-direction prompts for the content pipeline. Generate featured-image prompts, build stock image search queries, run the image-director reasoning that picks which sections get visuals (image_gen vs pexels, style, and prompt/query for each) plus one featured hero image, and caption rendered images with vision-derived alt text.
Generate a native, community-appropriate Reddit value-post from a published blog post
Blog drafting & revision prompts. Turn a topic + research context into a publication-ready markdown article, then refine it against reviewer feedback. Includes the writer system prompt, the draft request, the anti-fabrication initial-draft prompt, an SEO-metadata helper, and the iterative-refinement prompt. Use during the content-generation stage of the pipeline.
Video director — given a post body + podcast script + target duration, produces a JSON shot list (ordered shots with per-shot source plugin, prompt/query, and duration) for the post's video. Enforces the no-AI-humans + stylized-not-photoreal policies. Used by the generate_video_shot_list pipeline stage. Operator brand is templated via {site_name}.
Classify a published post into brand content-types (multi-label). Populates post_content_types via ClassifyContentTypesJob.
| name | hygiene |
| description | Data-retention and memory-hygiene summarization. Compress one calendar day of source-table rows, or a cluster of older same-source memories, into a single dense paragraph. Use when bounding storage growth while preserving the gist for retrieval. |
| license | Apache-2.0 |
| metadata | {"category":"utility","prompts":[{"key":"ops.retention.summarize_to_table","output_format":"text","description":"Compress one calendar day of rows from a source table into a single paragraph summary. Replaces N rows with 1 summary row to bound storage growth while preserving the gist for retrieval."},{"key":"memory.collapse_old_embeddings.summary","output_format":"text","description":"Compress N older same-source memory excerpts into a single dense paragraph for re-embedding. Future semantic search reads the summary instead of the originals — favor facts/names/decisions over prose."}]} |
Operational summarization prompts that bound storage growth.
ops.retention.summarize_to_table is used by
services/integrations/handlers/retention_summarize_to_table.py to
compress one calendar day of rows (e.g. audit_log) into a single summary
row. memory.collapse_old_embeddings.summary is consumed by
services/integrations/handlers/retention_embeddings_collapse.py (#242,
retired services/jobs/collapse_old_embeddings.py 2026-06-24) — a
per-policy-row config.prompt_template wins over the catalog default
(poindexter#829 wired the handler to this key).
UnifiedPromptManager resolves each template by key; both handlers
fetch the raw template through the gated _resolve_template_with_meta
seam and apply the {...} placeholders themselves. The inline fallbacks
in the handlers stay as a bootstrap safety net per
feedback_prompts_must_be_db_configurable and must stay byte-identical
to the bodies below (the shared drift guard enforces it).
You are compressing one calendar day of {source_table} rows so the system remembers the gist without storing every row. Below are {n} representative rows from {bucket_start_iso}, each separated by '---'.
Write a single paragraph (3-6 sentences) summarizing what happened that day. Preserve specific event types, sources, severity, decisions, and outcomes. Drop boilerplate and repetition. The summary will be stored as a single row replacing all {row_count} raw rows for the day, so dense factual content beats prose.
Rows:
{joined}
Summary:
You are compressing a cluster of older memories so the system remembers the gist without storing every detail. Below are {n} excerpts from the same source ({source_table}), each separated by '---'.
Write a single paragraph (3-6 sentences) summarizing what these excerpts collectively say. Preserve specific names, dates, decisions, errors, and outcomes. Drop boilerplate, repetition, and verbose phrasing. The summary will be embedded and used for future semantic search, so dense factual content beats prose.
Excerpts:
{joined}
Summary: