菜单
Session start context priming -- loads relevant memories from Pensyve at the beginning of a session to provide continuity across sessions. Use when switching projects or needing historical context.
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
Use when the user types @pensyve, asks for an @-mention workflow, wants mention-style Pensyve memory access in Codex, or asks whether Pensyve supports Codex mentions.
Use when the user explicitly invokes Pensyve, asks what the agent remembers, wants Codex to recall or store project memory, requests memory health review, or wants memory grounded work.
Pre-refactor context loading -- queries Pensyve memory for past decisions, failures, and patterns related to a refactoring target, then compiles a briefing. Use when starting a refactor to avoid repeating past mistakes.
End-of-session memory capture -- analyzes session for decisions, outcomes, and patterns worth remembering, then stores confirmed items via Pensyve. Use when ending a work session or when the user wants to capture what was learned.
Long-running multi-session work (research, eval loops, iterative benchmarks) with continuity -- resume prior lessons, capture per-run outcomes, build up stable truths over time. Use for eval/research/benchmark work that spans sessions.
Debug with working memory -- before diagnosing, recall prior root causes and known-good diagnostic procedures; when a root cause is confirmed, capture it immediately. Use whenever debugging a non-trivial failure.
| name | context-loader |
| description | Session start context priming -- loads relevant memories from Pensyve at the beginning of a session to provide continuity across sessions. Use when switching projects or needing historical context. |
| version | 1.0.0 |
Load relevant memories from Pensyve at the start of a session to provide cross-session continuity. Respects the context_loading configuration setting.
Note on invocation: The SessionStart hook has already performed a scoped recall and (if applicable) injected a continuity primer. Invoking this skill adds deeper on-demand context — not a replacement for the hook's work. Expect to see both.
When this skill is invoked (typically at session start), follow these steps:
Check the mode argument if provided. Otherwise, respect the context_loading setting from the plugin configuration (pensyve-plugin.local.md):
/recall to search memories on demand."If no mode is specified and no configuration is found, default to summary.
Run the following pensyve_recall queries to gather session context:
pensyve_recall with query "decided" (limit: 5)pensyve_recall with query "issue OR bug OR error OR problem" (limit: 5)pensyve_recall with query "workflow OR pattern OR process" (limit: 5)pensyve_recall with query "*" (limit: 10) -- broad query to capture recent memories by recencyDeduplicate results across queries (same memory ID should appear only once).
If the SessionStart hook detected a continuation (check plugin session state for recent_context — this is set by the hook when shared-entity score ≥0.7 against episodic recall results; it is a plugin-layer concept, not a structured server-side episode link), structure the primer around continuity rather than fresh recall:
Continuing:
<entity-set>— prior work contextLast lessons:
- [recent observation 1]
- [recent observation 2]
Open questions carried forward:
- [if any observations had
open-questionprovenance]
Note: Continuity is best-effort — it is inferred from shared-entity overlap in recent episodic recall, not from a persisted episode-to-episode link. The MCP server has no episode-listing API today. When in doubt, the hook falls back to a fresh-session primer.
If no continuation was detected, present a standard recall-based primer.
Present a concise briefing with the most important items:
Session Context (from Pensyve memory)
Recent Decisions:
- auth-service: Using RS256 for JWT signing to support key rotation
- api-design: POST endpoints return 201 with created resource
Known Issues:
- database: Migration script requires Python 3.11+
Active Patterns:
- testing: Integration tests need tmpdir cleanup after filesystem operations
Use
/recall <query>to search for specific memories.
Rules for summary mode:
Present a detailed briefing with scores and metadata:
Session Context (from Pensyve memory)
Recent Decisions (3 found):
Entity Decision Confidence When auth-service Using RS256 for JWT signing 0.9 2026-03-15 api-design POST endpoints return 201 0.9 2026-03-14 database SQLite for MVP, migrate to Postgres later 0.9 2026-03-12 Known Issues (2 found):
Entity Issue Confidence Score database Migration requires Python 3.11+ 0.8 0.91 cache Invalidation race condition on concurrent writes 0.8 0.85 Workflow Patterns (1 found):
Entity Pattern Confidence Score testing Integration tests need tmpdir cleanup 0.7 0.78 Recent Activity (5 unique memories, most recent first):
Type Entity Summary Score semantic auth-service RS256 JWT signing 0.92 episodic database Debugged migration failure 0.88 procedural deploy Run tests before deploy 0.75 Total memories loaded: 11 | Use
/recall <query>for targeted search |/inspect <entity>for entity details
Rules for full mode:
.claude/ memory files. All memory operations go through the Pensyve MCP tools exclusively./remember to start building context."pensyve_recall queries fail but others succeed, present the successful results and note the failures briefly.