parallel-memory
parallel-web/parallel-agent-skills
Recall past Parallel Task, Monitor, and FindAll runs to inform current work; manage memory with retrieve, evict, and clear operations.
What is parallel-memory?
Parallel Memory lets you search and recall results from prior Parallel Task, Monitor, and FindAll runs, helping you reuse past work when relevant. Use it to retrieve memories by topic or recency, evict individual runs, or clear your entire memory store. Requires parallel-cli >=0.8.1 and internet access.
- Retrieve past Parallel runs by semantic query or recency to find relevant prior work
- Filter retrievals by kind (task, monitor, findall) and timestamp boundaries
- Evict individual saved runs from your personal memory without deleting the underlying Parallel runs
- Clear your entire memory store with confirmation
- Access excerpts from task inputs/outputs, monitor events, and FindAll objectives
How to install parallel-memory
npx skills add https://github.com/parallel-web/parallel-agent-skills --skill parallel-memory- parallel-cli version >=0.8.1 installed and up to date
- Internet access
- Memory eligibility (rollout, organization settings, account opt-in, and API key eligibility apply)
How to use parallel-memory
- 1.Run `parallel-cli memory retrieve --query "<topic>"` to search memory by semantic query, or omit query and add `--limit N` to see recent runs
- 2.Review the returned excerpts (id, updated_at, input_excerpt, output_excerpt for tasks; status and event details for monitors; objective and matched_count for findalls)
- 3.Fetch the full original run result when you need exact output, entities, or citations using the returned run ID
- 4.Run `parallel-cli memory evict --kind <task|monitor|findall> --id "<run-id>"` to remove a single run from memory
- 5.Run `parallel-cli memory clear --confirm-clear` to permanently remove all memory entries
Use cases
- Search memory for prior work on a topic (e.g., 'serverless inference vendors') to avoid redundant analysis
- Review recent past runs to understand what work has already been completed
- Remove a specific saved run from memory when it is no longer relevant
- Permanently clear all memory entries before sharing access or resetting your workspace
- Fetch full details from a recalled run when you need exact output, citations, or provenance
- Parallel users managing repeated or iterative analysis tasks
- Teams working on projects where prior research and runs should inform new work
- Users who want to avoid duplicate effort by recalling past Parallel executions
parallel-memory FAQ
You get excerpts from past Parallel Task, Monitor, and FindAll runs matching your query or recency filter. Empty results means no matches were found, not an error. Always fetch the original run for full details, citations, and provenance.
No. Memory retrieval is separate from launching new Parallel runs. Parallel runs do not consult Memory automatically; if recalled information may help, include relevant details in the new run's input.
No. Ingestion is asynchronous, so newly completed runs may not be immediately retrievable. Wait a moment before expecting a fresh run to appear in memory results.
Evict removes a single saved run by kind and ID. Clear removes all entries from your personal Memory. Neither deletes the underlying Parallel runs. Clear requires confirmation unless the user already asked for it.
Reauthenticate with `parallel-cli`. If memory is not eligible for other reasons (rollout, organization settings, account opt-in), the error message will explain the restriction.
Full instructions (SKILL.md)
Source of truth, from parallel-web/parallel-agent-skills.
name: parallel-memory description: "Recall past Parallel Task, Monitor, and FindAll runs when they may help; evict runs or clear memory when asked." user-invocable: true argument-hint: <retrieve|evict|clear> [args] compatibility: Requires parallel-cli >=0.8.1 and internet access. allowed-tools: Bash(parallel-cli:*) metadata: author: parallel
Parallel Memory
Action: $ARGUMENTS
Requires
parallel-cli >=0.8.1. If the installed version is older, orparallel-cli memory --helpfails withno such commandor similar, tell the user to updateparallel-cli, then retry.
When to use
- Inspect memory only when prior Parallel work may help fulfill the request or the user asks to retrieve, evict, or clear it.
- Memory results are excerpts from past runs; fetch the source run for full records, and launch a fresh run when current information is required.
Choose the operation
| User intent | Operation |
|---|---|
| Recall prior work about a topic | Retrieve with a concise query |
| Show recent past runs | Retrieve without query |
| Remove one saved Task, Monitor, or FindAll source | Evict by exact kind and id |
| Permanently remove all entries from your personal Memory | Clear memory |
| Turn memory off | Direct the user to account settings; do not clear as a substitute |
Use the CLI
Use parallel-cli memory for retrieve, evict, and clear operations.
- If memory is not eligible, report the returned reason; it distinguishes rollout, organization settings, account opt-in, and key eligibility.
- On a key-eligibility error, tell the user to reauthenticate.
Retrieve memory
Form a short semantic query that describes the prior work to find. Apply filters when they help. Empty results is a successful retrieval with no matches, not an error.
- Set
kindtotask,monitor, orfindallwhen it clearly narrows the retrieval. - Set
sincefor an explicit timestamp boundary (RFC 3339, e.g.2026-08-01T00:00:00Z). - Omit
querywhen retrieving recent memories rather than a topic.
Retrieve by query:
parallel-cli memory retrieve \
--query "serverless inference vendors"
For recent memories:
parallel-cli memory retrieve \
--limit 5
Use results
Available fields vary by kind:
-
task: useid,updated_at,input_excerpt, andoutput_excerpt. -
monitor: use the monitorid, status, query excerpt, and matching event IDs, timestamps, and excerpts. -
findall: useid,updated_at, objective excerpt, andmatched_count. -
Fetch the original Task result, Monitor events, or FindAll result when exact output, entities, citations, or provenance matter.
-
Summarize the useful findings and unresolved questions.
-
Lead with what the prior work established, then list the contributing saved runs with kind, ID, and timestamp.
-
Distinguish recalled information from any fresh verification.
-
Parallel runs do not consult Memory. If recalled information may be useful, include the relevant details in the new run's input.
Expect ingestion to be asynchronous. Do not promise that a newly completed run will be immediately retrievable.
Evict or clear memory
Evict a single run from your personal Memory, or clear it entirely. These do not delete the underlying Parallel runs. Ask for confirmation before clearing unless the user already asked for it.
parallel-cli memory evict \
--kind task \
--id "trun_example"
parallel-cli memory clear \
--confirm-clear
Related skills
More from parallel-web/parallel-agent-skills and the wider catalog.

parallel-monitor
Continuously monitor web pages and track changes on a recurring schedule with server-side event detection.

parallel-web-extract
CLI-backed URL extraction with JSON output when web_fetch MCP is unavailable.

parallel-web-search
CLI-backed web search with JSON output and keyword filtering.

result
Retrieve completed research task results by run ID using Parallel CLI.

setup
Install and authenticate the Parallel CLI for agent integration.

status
Check the status of a running research task by its run ID.