Skip to content
Skills Library
Management & product · Planning

codex-project-memory

/codex-project-memory

Use for project memory, decisions, handoffs, genome, knowledge index, changelog, and patterns that must persist across sessions.

Licence
MIT
Last repository change
23.10.2026

Assessment and security

100/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
36
Scripts
15
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 23.10.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What to check before installing
  • 15 scripts in this package. Review them before installing, as Claude may execute them.

Scripts: scripts/analyze_patterns.py, scripts/build_knowledge_graph.py, scripts/build_knowledge_index.py, scripts/codebase_indexer.py, scripts/compact_context.py, scripts/decision_logger.py, scripts/generate_changelog.py, scripts/generate_genome.py

What it does and when to use it

Use for project memory, decisions, handoffs, genome, knowledge index, changelog, and patterns that must persist across sessions.

Original author description

Use for project memory, decisions, handoffs, genome, knowledge index, changelog, and patterns that must persist across sessions.

How it works

  1. TL;DR
  2. Activation
  3. Decision Tree Routing
  4. Rules
  5. Reference Files
Original SKILL.md headings · 83 lines, 655 words
SKILL.md preview
---
name: codex-project-memory
description: Use for project memory, decisions, handoffs, genome, knowledge index, changelog, and patterns that must persist across sessions.
load_priority: on-demand
---

## TL;DR
10 scripts for project knowledge persistence. Decision tree: Log -> decision/feedback. Generate -> changelog/session-summary/handoff. Analyze -> patterns/knowledge-graph/knowledge-index/project-genome. Maintain -> compact old context. Triggers: `$changelog`, `$session-summary`, `$log-decision`, `$compact-context`, `$knowledge`, `$codex-genome`.

# Codex Project Memory

## Activation

1. Activate after complex tasks involving architecture, design decisions, or refactors.
2. Activate on explicit `$codex-project-memory` or `$memory`.
3. Activate on explicit command `$log-decision`.
4. Activate on `$handoff` or when user asks to generate handoff context.
5. Activate on `$session-summary` or when user asks to summarize a coding session.
6. Activate on `$analyze-patterns` or "learn project style".
7. Activate on `$log-feedback` or "track my fix".
8. Activate on `$feedback-summary` for aggregated feedback recall.
9. Activate on `$build-graph` or "map project" to generate project knowledge graph.
10. Activate on `$changelog` to generate release-oriented commit summaries.
11. Activate on `$compact-context` or "clean up old sessions".
12. Activate on `$codex-genome`, `$generate-genome`, or `$genome` to create layered project context docs.
13. Activate on `$knowledge`, "make tacit knowledge visible", or "build knowledge index" to generate `.codex/knowledge/INDEX.md`, `.codex/knowledge/index.json`, and `.codex/knowledge/knowledge-graph.json`.
14. Activate on "memory status", "validate project memory", or before CLI/tool-harness handoff to run `scripts/memory_status.py` against generated artifacts.
15. Before pausing multi-step work for context compaction or handoff, create a session-summary checkpoint and complete its task-state fields from the current session.

## Decision Tree Routing

```
User intent -> Memory action?
    |- Log/record -> What?
    |   |- Decision -> decision_logger.py
    |   `- Feedback -> track_feedback.py
    |
    |- Generate/export -> What?
    |   |- Handoff -> generate_handoff.py
    |   |- Session recap -> generate_session_summary.py
    |   `- Changelog -> generate_changelog.py
    |
    |- Analyze -> What?
    |   |- Patterns -> analyze_patterns.py
    |   |- Dependencies -> build_knowledge_graph.py
    |   |- Tacit knowledge index -> build_knowledge_index.py
    |   `- Context genome -> generate_genome.py
    |
    |- Maintain -> What?
    |   `- Compact old memory -> compact_context.py
    |
    `- Not memory -> skip
```

## Rules

- Follow `references/unified-memory-architecture.md` for the 4-Tier Memory Hierarchy (L0 Execution State -> L1 Genome/Knowledge Index -> L2 Normative Specs/ADRs -> L3 Role Docs), SSOT conflict precedence, and deterministic Write-Back Protocol.
- Use the decision tree to choose one memory action before invoking scripts.
- Run `--help` before invoking helper scripts and treat them as black-box tools unless customization or bug fixing is required.
- Review existing files in `.codex/decisions/` before logging a similar technical choice.
- Use the default output paths and chaining guidance from `references/behavior-guide.md` unless the user requests overrides.
- A generated session summary is a Git snapshot, not a complete task handoff. Fill its task-state, active spec/ACs, SILC integration seam status, verification, and next-step prompts before treating it as resumable memory.
- `--since last-session` is scoped to the current repository, branch, and reachable commit history; complete the newest matching checkpoint before using it as the next boundary. It is a Git-history selector, not a task selector.
- Return the created path plus the headline metrics from each script JSON payload.
- See `skills/.system/REGISTRY.md` for full script paths.

## Reference Files

- `references/unified-memory-architecture.md`: 4-tier SSOT read hierarchy, authority precedence, and deterministic write-back protocol across sessions.
- `references/decision-journal-spec.md`: decision logging criteria and naming conventions.
- `references/handoff-spec.md`: handoff usage guidance.
- `references/session-summary-spec.md`: session summary workflow and retention guidance.
- `references/changelog-spec.md`: release-focused changelog generation rules.
- `references/pattern-learner-spec.md`: project style learning behavior.
- `references/feedback-tracker-spec.md`: feedback logging and aggregate usage.
- `references/knowledge-graph-spec.md`: deep architecture mapping and refresh guidance.
- `build_knowledge_index.py`: schema-versioned project knowledge index from genome, role docs, decisions, commits, config, and the structural knowledge graph; redacts secret-like values, stores provenance/confidence and a source fingerprint for staleness checks.
- `references/knowledge-index.schema.json`: output contract for `.codex/knowledge/index.json`.
- `references/knowledge-graph.schema.json`: output contract for `.codex/knowledge/knowledge-graph.json` and `.codex/knowledge-graph.json`.
- `references/context-compactor-spec.md`: retention policy, archive layout, and dry-run expectations.
- `references/behavior-guide.md`: detailed per-script behaviors, defaults, and chaining guidance.
- `references/script-commands.md`
- `references/output-schemas.md`
- `references/project-memory-tools.json`: machine-readable tool manifest for CLI/MCP-style harness generation.
- `references/project-memory-tools.schema.json`: schema for the tool manifest.
- `references/ci-readiness.md`: release gate commands, strict mode, and Windows symlink deselection.
- `references/artifact-lifecycle-policy.md`: hybrid commit policy for schemas vs generated `.codex/` artifacts.
Files in this package (36)
  • SKILL.md 6 KB
  • agents/openai.yaml 1 KB
  • references/artifact-lifecycle-policy.md 4 KB
  • references/behavior-guide.md 7 KB
  • references/changelog-spec.md 1 KB
  • references/ci-readiness.md 5 KB
  • references/codebase-index.schema.json 2 KB
  • references/context-compactor-spec.md 1 KB
  • references/decision-journal-spec.md 1 KB
  • references/feedback-tracker-spec.md 3 KB
  • references/handoff-spec.md 6 KB
  • references/knowledge-graph-spec.md 6 KB
  • references/knowledge-graph.schema.json 4 KB
  • references/knowledge-index.schema.json 3 KB
  • references/output-schemas.md 11 KB
  • references/pattern-learner-spec.md 4 KB
  • references/project-memory-tools.json 7 KB
  • references/project-memory-tools.schema.json 2 KB
  • references/script-commands.md 7 KB
  • references/session-summary-spec.md 7 KB
  • references/unified-memory-architecture.md 6 KB
  • scripts/analyze_patterns.py 28 KB
  • scripts/build_knowledge_graph.py 58 KB
  • scripts/build_knowledge_index.py 29 KB
  • scripts/codebase_indexer.py 23 KB
  • scripts/compact_context.py 13 KB
  • scripts/decision_logger.py 5 KB
  • scripts/generate_changelog.py 8 KB
  • scripts/generate_genome.py 3 KB
  • scripts/generate_handoff.py 22 KB
  • scripts/generate_session_summary.py 25 KB
  • scripts/genome_builder.py 44 KB
  • scripts/memory_status.py 15 KB
  • scripts/project_traversal.py 18 KB
  • scripts/redaction.py 4 KB
  • scripts/track_feedback.py 11 KB

Versions

The database has no other skills named codex-project-memory.

Database snapshot: 13.11.2026 · package version ba70a343c05d (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 90project-memory
    Use when working in a repository that keeps durable project context, cross-session memory, incident history, decisions, or agent state across Claude, Codex, and later sessions.
  • 98shared-project-memory
    开始或继续本地项目任务时读取 Codex 与 Claude 共用的项目记忆;完成任务阶段、遇到阻塞或准备切换工具时保存交接;查询以前的项目决策与排障记录。使用本机文件和关键词检索,无额外模型 API。适用于本机 Code 会话;普通问答和无项目目录的聊天无需使用。
  • 90project-context
    Restore durable PowerContext project memory and transfer work between OpenCode sessions. Use when continuing prior work, recalling decisions or constraints, preparing a handoff, or explicitly maintaining project Memory.
  • 90project-context
    Restore durable PowerContext project memory and transfer work between Pi sessions. Use when continuing prior work, recalling a decision or constraint, preparing a handoff, or explicitly maintaining project Memory.
  • 96context-memory-keeper
    维护跨会话的项目记忆,记录长期规则与当前任务清单,供后续阶段恢复上下文。当长对话需要压缩、已确认的决策与约束需要落盘、或用户要求记住或恢复之前进展时使用。
  • 90briefbound-project-memory
    Use when the user or project rules request project-local memory/dashboard initialization or updates, or an existing `.docs/project-memory/` or resolved external `activePaths.memory` root is needed for cross-session recovery, durable decisions, blockers, parallel coordination, or…

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.