+1
-1
@@ -9,2 +9,2 @@ /** | ||
| */ | ||
| export const VERSION = "0.10.3"; | ||
| export const VERSION = "0.10.4"; |
+1
-1
| { | ||
| "name": "jamgate", | ||
| "version": "0.10.3", | ||
| "version": "0.10.4", | ||
| "mcpName": "io.github.amirj4m/jamgate", | ||
@@ -5,0 +5,0 @@ "description": "A neutral, cross-agent memory quality gate for AI agents, delivered as an MCP server — a gate, not a store.", |
+24
-5
@@ -243,2 +243,11 @@ # Jamgate | ||
| Volatile memories age out of recall on a TTL, so a fact you saved can stop being returned | ||
| without being deleted. `jamgate expired` shows exactly what recall is hiding and when it will | ||
| be compacted away — nothing is modified: | ||
| ```bash | ||
| jamgate expired # what has aged out of recall but is still on disk | ||
| jamgate expired --json # machine-readable, for a script | ||
| ``` | ||
| Your memory lives in `~/.jamgate/memory.json`. Same machine, every agent → one shared | ||
@@ -257,4 +266,5 @@ memory. To share one memory across **different** machines and your phone, see | ||
| It ships in this repo at [`skills/memory-discipline/SKILL.md`](./skills/memory-discipline/SKILL.md) | ||
| as a portable [agentskills.io](https://agentskills.io) instruction pack, installable into | ||
| 70+ coding agents with one command: | ||
| as a portable [agentskills.io](https://agentskills.io) instruction pack. One command installs it | ||
| for every agent the `skills` CLI finds on your machine (it wired 17, including Cursor, Copilot | ||
| and Claude Code, on the machine we tested): | ||
@@ -768,2 +778,7 @@ ```bash | ||
| The Jamgate column is verified by the test suite. The other two columns are read from those | ||
| projects' own public documentation as of **August 2026** and describe their default behaviour, | ||
| not the ceiling of what they can be configured to do — if we have a row wrong, open an issue | ||
| and we'll fix it. | ||
| Mem0, OpenMemory, Zep, and Graphiti are capable systems built for different goals; if you | ||
@@ -777,4 +792,8 @@ need managed scale or graph reasoning, they're the right tool. Jamgate's bet is that for | ||
| - **Everything is local.** The memory store, the gate, and (if enabled) the embedding | ||
| model all run on your machine. Jamgate makes **no network calls** and talks to no cloud | ||
| AI. | ||
| model all run on your machine. **Your memories are never sent anywhere** — Jamgate makes | ||
| no outbound request carrying your data, has no telemetry, and talks to no cloud AI. | ||
| The one network access in the whole product is the optional embedding model's **first-run | ||
| download** from Hugging Face (~23 MB, cached, and only if you installed that optional | ||
| package); it downloads weights, it does not upload text. `npx` itself fetching the package | ||
| from npm is the other, and both stop after install. | ||
| - **The decision log is local too.** Every gate decision (saved / duplicate / superseded / | ||
@@ -821,3 +840,3 @@ conflict / possible_duplicate / rejected, with its reason) is appended to a strictly | ||
| Verified end-to-end over the MCP protocol (both stdio and HTTP) and covered by an automated | ||
| test suite (413 tests) on Node 20.x and 22.x. Next: a thin classifier for ambiguous cases | ||
| test suite (460+ tests) on Node 20.x and 22.x. Next: a thin classifier for ambiguous cases | ||
| (trained on the local decision log) and multi-device sync (see [`DECISIONS.md`](./DECISIONS.md)). | ||
@@ -824,0 +843,0 @@ **Goal: impact, not profit — open-source (MIT), built in the open.** |
URL strings
Supply chain riskPackage contains fragments of external URLs or IP addresses, which the package may be accessing at runtime.
URL strings
Supply chain riskPackage contains fragments of external URLs or IP addresses, which the package may be accessing at runtime.
316036
0.38%860
2.26%