How do I install the Aim Write skill?

Run npx skillmds@latest add djalmajr/aim-write in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Aim Write safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Aim Write?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Aim Write free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Aim Write?

djalmajr (@djalmajr) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Aim Write by djalmajr — Agent Skill

Name: Aim Write
Author: djalmajr

Aim Write

Manually write a durable page to an ai-memory instance via the MCP (memory_write_page) — for decisions, rules, gotchas, schemas/contracts, or operational constraints the user explicitly wants remembered. Use when the user says 'remember this', 'save this decision', or 'annotate X'. Not for routine notes (auto-capture handles those).

djalmajr 7dbfc33 2.6 KB Updated Sep 17, 2026

File contents

aim-write

Persist a durable wiki page. Use only for canonical knowledge the user explicitly wants remembered — routine session notes are captured automatically by the hooks; do not hand-write those, and do not use a handoff for a permanent annotation.

Steps

Pick the instance. From the repo's MCP config. If multiple ai-memory servers are configured, ask which endpoint to write to.
Pick workspace/project. Default to the repo's .ai-memory.toml; confirm if writing elsewhere.
Choose a path + kind. Use folder conventions so kind is derived and recall stays clean:
- decisions/ADR-NNNN-<slug>.md → decision
- _rules/<slug>.md → rule
- gotchas/<slug>.md → gotcha
- otherwise a topical path (e.g. runbooks/<slug>.md, <area>/<slug>.md). You can also set kind explicitly via frontmatter node_type / the tool arg.
The _rules/ prefix matters — use the underscore. Pages under _rules/ (and pinned _slots/) are ai-memory's highest-signal tier: they're surfaced verbatim in memory_briefing / memory_explore / the dashboard, keyed by path GLOB '_rules/*'. A plain rules/ folder (no underscore) is just a normal page — it gets none of that and is only found by explicit query/read. So write durable rules to _rules/, never rules/.

Shared / cross-project rules (e.g. global agent conventions reused across every repo) belong in a dedicated scope — the reserved default/_global scope — under _rules/ there (write with scope: "global"). Recall is per-(workspace, project), so other projects reach them cross-scope via memory_query scopes:[…] / memory_read_page (see aim-query), not automatically.
Write the body as markdown with frontmatter (title, node_type, tags, optional source:). Link related pages with [[path.md]] (paths carry .md; links resolve by exact path). Never put live secrets/credentials in a page — redact (<see secret manager>).
Call memory_write_page (or the equivalent admin write) with workspace/project/path/body.
Confirm the page is recallable (a quick aim-query for its topic).

Keep pages focused and distilled — one fact/decision/gotcha per page reads and recalls best.

---name: aim-writedescription: Manually write a durable page to an ai-memory instance via the MCP (memory_write_page) — for decisions, rules, gotchas, schemas/contracts, or operational constraints the user explicitly wants remembered. Use when the user says 'remember this', 'save this decision', or 'annotate X'. Not for routine notes (auto-capture handles those).---# aim-writePersist a durable wiki page. Use only for **canonical** knowledge the user explicitly wantsremembered — routine session notes are captured automatically by the hooks; do not hand-writethose, and do not use a handoff for a permanent annotation.## Steps1. **Pick the instance.** From the repo's MCP config. If multiple ai-memory servers are configured, ask which endpoint to write to.2. **Pick workspace/project.** Default to the repo's `.ai-memory.toml`; confirm if writing elsewhere.3. **Choose a path + kind.** Use folder conventions so `kind` is derived and recall stays clean: - `decisions/ADR-NNNN-<slug>.md` → decision - `_rules/<slug>.md` → rule - `gotchas/<slug>.md` → gotcha - otherwise a topical path (e.g. `runbooks/<slug>.md`, `<area>/<slug>.md`). You can also set `kind` explicitly via frontmatter `node_type` / the tool arg. **The `_rules/` prefix matters — use the underscore.** Pages under `_rules/` (and pinned `_slots/`) are ai-memory's **highest-signal tier**: they're surfaced **verbatim** in `memory_briefing` / `memory_explore` / the dashboard, keyed by `path GLOB '_rules/*'`. A plain `rules/` folder (no underscore) is just a normal page — it gets **none** of that and is only found by explicit query/read. So write durable rules to `_rules/`, never `rules/`. **Shared / cross-project rules** (e.g. global agent conventions reused across every repo) belong in a dedicated scope — the reserved `default`/`_global` scope — under `_rules/` there (write with `scope: "global"`). Recall is per-(workspace, project), so other projects reach them **cross-scope** via `memory_query scopes:[…]` / `memory_read_page` (see `aim-query`), not automatically.4. **Write the body** as markdown with frontmatter (`title`, `node_type`, `tags`, optional `source:`). Link related pages with `[[path.md]]` (paths carry `.md`; links resolve by exact path). **Never put live secrets/credentials in a page** — redact (`<see secret manager>`).5. **Call `memory_write_page`** (or the equivalent admin write) with workspace/project/path/body.6. **Confirm** the page is recallable (a quick `aim-query` for its topic).Keep pages focused and distilled — one fact/decision/gotcha per page reads and recalls best.

Frequently asked questions