# AI use cases for Shared Memory Vault (songs-aaa/shared-memory-vault) Repo: [songs-aaa/shared-memory-vault](https://github.com/songs-aaa/shared-memory-vault) | Use cases: 5 | Page: https://dedupedai.com/use-cases/songs-aaa-shared-memory-vault/ A Markdown-based shared memory and governance toolkit that lets multiple AI agents and devices read and write one cloud-synced knowledge base. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Set up a shared knowledge base in a cloud folder: https://dedupedai.com/use-cases/songs-aaa-shared-memory-vault/bootstrap-shared-knowledge-base/ - Hand off an unfinished task from one agent to another: https://dedupedai.com/use-cases/songs-aaa-shared-memory-vault/hand-off-work-between-agents/ - Score a knowledge base for governance and provenance: https://dedupedai.com/use-cases/songs-aaa-shared-memory-vault/audit-knowledge-base-governance/ - Find earlier decisions before starting new work: https://dedupedai.com/use-cases/songs-aaa-shared-memory-vault/search-past-project-memory/ - Distill reflections and link memories across projects: https://dedupedai.com/use-cases/songs-aaa-shared-memory-vault/reflect-and-link-project-memory/ ## How to set up a shared knowledge base in a cloud folder with Shared Memory Vault Repo: [songs-aaa/shared-memory-vault](https://github.com/songs-aaa/shared-memory-vault) | Difficulty: Intermediate | Time: 15 minutes | Basis: From the README Create a Markdown knowledge base skeleton that every agent and device can read from one synced folder. **Who it is for:** A developer with several AI agents and two or more devices who wants one place for shared project memory that syncs through an existing cloud folder. **The problem:** Each agent and device keeps its own notes, so context drifts and decisions get repeated or contradicted. **The outcome:** A single synced directory with the standard zones (protocol, project memory, session blackboard, references, templates, skills and tools) that agents follow through SKILL.md. ### Steps 1. Copy the skill folder into your skills location: `cp -r shared-memory-vault ~/.workbuddy/skills/` 2. Create the knowledge base in your cloud-synced directory: `bash ~/.workbuddy/skills/shared-memory-vault/scripts/vault-init.sh --dir "~/MyCloud/KnowledgeBase"` 3. Open the new folder and confirm it contains MOC.md, dashboard.md, user memory/, work memory/, context/, references/, _templates/ and skill&tool/. 4. Point each agent at SKILL.md so it follows the lifecycle SOP automatically. ### Prompt to paste into your agent ```text Using the shared-memory-vault repo, bootstrap a knowledge base skeleton at by running its vault-init.sh script with --dir. Report the top-level entries created. Verify that MOC.md and context/ exist before you finish. ``` ### Caveat The README gives only the --dir flag for vault-init.sh, so check the script itself for any other options before relying on them. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/songs-aaa-shared-memory-vault/bootstrap-shared-knowledge-base/ ## How to hand off an unfinished task from one agent to another with Shared Memory Vault Repo: [songs-aaa/shared-memory-vault](https://github.com/songs-aaa/shared-memory-vault) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Let a second agent resume work from a shared CONTEXT.md without re-explaining the task. **Who it is for:** A developer whose coding agent hits a session limit or switches device and needs a second agent to continue the same task. **The problem:** The next agent starts cold, repeats earlier steps, or makes decisions that conflict with the first agent's choices. **The outcome:** The second agent reads one CONTEXT.md file and continues from where the first agent stopped, with decisions kept consistent. ### Steps 1. Set up the knowledge base once with vault-init.sh so both agents share the same folder. 2. Have the first agent record its goal, decisions and next step in CONTEXT.md for the session. 3. Have agents update CONTEXT.md incrementally every 3 turns: `python3 ~/.workbuddy/skills/shared-memory-vault/scripts/context-update.py` (inferred, check the README) 4. Start the second agent and tell it to read CONTEXT.md first, then continue the task. 5. When the task pauses, have the agent archive the session with vault-archive.py so the next handoff starts from a clean blackboard. ### Prompt to paste into your agent ```text Using the shared-memory-vault repo, read the CONTEXT.md file in /context/ and resume the task it describes. Report the next step you take and the decision you are following from the file. Verify that you did not change any earlier decision without noting it. ``` ### Caveat The README describes the blackboard convention but does not define the exact CONTEXT.md fields, so agree on a format with your agents before the first handoff. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/songs-aaa-shared-memory-vault/hand-off-work-between-agents/ ## How to score a knowledge base for governance and provenance with Shared Memory Vault Repo: [songs-aaa/shared-memory-vault](https://github.com/songs-aaa/shared-memory-vault) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Check tagging and governance quality of shared notes and get a periodic score to fix gaps. **Who it is for:** A team lead or maintainer who relies on agent-written notes and wants to confirm they are tagged and governed before trusting them. **The problem:** Agent-written notes accumulate without clear sources or confidence levels, so nobody knows which entries to trust. **The outcome:** A provenance check and a governance benchmark that list where notes lack source or confidence tags and give an overall score. ### Steps 1. Run the provenance check on the knowledge base: `python3 ~/.workbuddy/skills/shared-memory-vault/scripts/provenance-check.py` (inferred, check the README) 2. Read the list of notes flagged for missing provenance or confidence tags. 3. Run the governance benchmark: `python3 ~/.workbuddy/skills/shared-memory-vault/scripts/governance-score.py` (inferred, check the README) 4. Fix the flagged notes, then rerun the benchmark to confirm the score improves. ### Prompt to paste into your agent ```text Using the shared-memory-vault repo, run its provenance check and governance benchmark on the knowledge base at . Report the governance score and the number of notes flagged for provenance. Verify the score by rerunning the benchmark after any fixes. ``` ### Caveat The README names these scripts but does not show their command-line arguments, so the commands above are inferred and should be checked against the script headers. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/songs-aaa-shared-memory-vault/audit-knowledge-base-governance/ ## How to find earlier decisions before starting new work with Shared Memory Vault Repo: [songs-aaa/shared-memory-vault](https://github.com/songs-aaa/shared-memory-vault) | Difficulty: Intermediate | Time: 15 minutes | Basis: From the README Query the shared knowledge base so an agent reuses past decisions instead of redoing research. **Who it is for:** A developer starting a new task in a project that has months of logs, notes and reflections already in the shared vault. **The problem:** Past decisions and research sit in many files, so agents or people redo work or contradict an earlier choice. **The outcome:** A ranked set of relevant past notes and reflections that the agent reads before it plans the new task. ### Steps 1. Make sure the knowledge base exists and contains project memory under work memory/. 2. Run a semantic search with your question: `python3 ~/.workbuddy/skills/shared-memory-vault/scripts/search.py ""` (inferred, check the README) 3. Open the top results and note any decisions that apply to the new task. 4. Have the agent cite those notes in its plan before it starts work. ### Prompt to paste into your agent ```text Using the shared-memory-vault repo, run its semantic search for against the knowledge base at . Report the three most relevant notes and one decision from each. Verify that each decision points to a real file in the vault. ``` ### Caveat The README lists search.py under semantic retrieval but does not describe how it takes its query, so confirm the argument format in the script before using the command above. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/songs-aaa-shared-memory-vault/search-past-project-memory/ ## How to distill reflections and link memories across projects with Shared Memory Vault Repo: [songs-aaa/shared-memory-vault](https://github.com/songs-aaa/shared-memory-vault) | Difficulty: Advanced | Time: 45 minutes | Basis: From the README Generate periodic reflections and connect related notes from different projects in the vault. **Who it is for:** A researcher or developer running several projects in one knowledge base who wants lessons from one project to surface in the others. **The problem:** Lessons learned in one project stay buried, and related notes across projects never reference each other. **The outcome:** Reflection notes summarising recent work and cross-project links between related memories. ### Steps 1. Generate reflections from recent project memory: `python3 ~/.workbuddy/skills/shared-memory-vault/scripts/reflect.py` (inferred, check the README) 2. Review the new reflection notes in the project memory area. 3. Build cross-project links between memories: `python3 ~/.workbuddy/skills/shared-memory-vault/scripts/link-memory.py` (inferred, check the README) 4. Open the dashboard or MOC.md to confirm the new links appear in navigation. ### Prompt to paste into your agent ```text Using the shared-memory-vault repo, run its reflection and cross-project linking scripts on the knowledge base at . Report how many reflection notes were created and how many links were added. Verify that each new link points to an existing note. ``` ### Caveat The README places reflection and linking on a milestone or schedule, but it does not say how those triggers are set up, so schedule them yourself and check the script output. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/songs-aaa-shared-memory-vault/reflect-and-link-project-memory/