fix(context): strip .md suffix using correct byte length

The .md extension is three bytes, but the normalizer took the last four
characters as the suffix, so USER.md compared R.md to .md and never
stripped. Use MD_SUFFIX.len() for slicing and document optional .md on
show/pull/push in help and the hotdata skill.

Author: eddietejeda

skills/hotdata/SKILL.md

@@ -43,7 +43,7 @@ If **`HOTDATA_WORKSPACE`** is set in the environment, the workspace is **locked*
 
 The workspace stores those documents only through the **context API**. The **authoritative** copy always lives on the server under the stem; common stems are **`context:DATAMODEL`** (semantic map) and **`context:GLOSSARY`** (glossary / runbooks).
 
-The CLI command **`hotdata context push`** reads **`./<NAME>.md`** and **`pull`** writes that file in the **current working directory**—those files exist only as a **transport surface** for the API, not as a second source of truth. **`hotdata context show <name>`** prints Markdown to stdout so agents can read **`context:<NAME>`** without any local file. Stems follow SQL table–identifier rules (ASCII letters, digits, underscore; no dot in the API name; max 128 characters; SQL reserved words are not allowed).
+The CLI command **`hotdata context push`** reads **`./<NAME>.md`** and **`pull`** writes that file in the **current working directory**—those files exist only as a **transport surface** for the API, not as a second source of truth. **`hotdata context show <name>`** prints Markdown to stdout so agents can read **`context:<NAME>`** without any local file. Stems follow SQL table–identifier rules (ASCII letters, digits, underscore; no dot in the API name; max 128 characters; SQL reserved words are not allowed). For **`show`**, **`pull`**, and **`push`**, the CLI accepts a trailing **`.md`** on the argument (e.g. **`USER.md`**) and treats it as stem **`USER`**—the workspace still stores **`USER`**, not `USER.md`.
 
 > **Agents: do not blindly run `hotdata context show DATAMODEL` on session start.** Run **`hotdata context list`** first (optional `--prefix DATAMODEL`). Call **`hotdata context show DATAMODEL` only if** the list includes the `DATAMODEL` stem. If **`show` exits 1** with *no context named …*, that is **normal** when nothing has been pushed yet—**not a hard failure**; do not retry in a loop, and **avoid speculative `show` in parallel** with other shell tools where one failure cancels sibling calls. Proceed without **context:DATAMODEL** until the user asks to create or load one.
 

src/command.rs

@@ -582,13 +582,13 @@ pub enum ContextCommands {
 
     /// Print context content to stdout
     Show {
-        /// Context name (same rules as a SQL table identifier; local file is <NAME>.md)
+        /// Context name (same rules as a SQL table identifier; local file is <NAME>.md). A trailing `.md` is ignored (e.g. `USER.md` → `USER`).
         name: String,
     },
 
     /// Download context from the workspace to ./<NAME>.md
     Pull {
-        /// Context name
+        /// Context name (trailing `.md` ignored, e.g. `USER.md` → `USER`)
         name: String,
 
         /// Overwrite ./<NAME>.md if it already exists
@@ -602,7 +602,7 @@ pub enum ContextCommands {
 
     /// Upload ./<NAME>.md to the workspace as named context
     Push {
-        /// Context name
+        /// Context name (trailing `.md` ignored, e.g. `USER.md` → `USER`; reads `./USER.md`)
         name: String,
 
         /// Print what would be sent; do not POST

src/context.rs

@@ -49,6 +49,25 @@ struct UpsertResponse {
     context: WorkspaceContextEntry,
 }
 
+/// Normalizes a context name from the CLI: trims, takes the final path segment, and strips a
+/// trailing `.md` (any ASCII case) so `USER.md` or `./USER.md` refer to context stem `USER`.
+pub fn normalize_context_cli_name(name: &str) -> String {
+    let trimmed = name.trim();
+    let basename = std::path::Path::new(trimmed)
+        .file_name()
+        .and_then(|n| n.to_str())
+        .unwrap_or(trimmed);
+    const MD_SUFFIX: &str = ".md";
+    if basename.len() >= MD_SUFFIX.len() {
+        let start = basename.len() - MD_SUFFIX.len();
+        let suffix = &basename[start..];
+        if suffix.eq_ignore_ascii_case(MD_SUFFIX) {
+            return basename[..start].to_string();
+        }
+    }
+    basename.to_string()
+}
+
 /// Validates a context stem (API `name` and basename before `.md`).
 /// Same rules as runtimedb `validate_table_name`.
 pub fn validate_context_stem(name: &str) -> Result<(), String> {
@@ -148,13 +167,14 @@ pub fn list(workspace_id: &str, prefix: Option<&str>, format: &str) {
 }
 
 pub fn show(workspace_id: &str, name: &str) {
-    if let Err(e) = validate_context_stem(name) {
+    let name = normalize_context_cli_name(name);
+    if let Err(e) = validate_context_stem(&name) {
         eprintln!("error: {e}");
         std::process::exit(1);
     }
 
     let api = ApiClient::new(Some(workspace_id));
-    match fetch_context(&api, name) {
+    match fetch_context(&api, &name) {
         Ok(ctx) => {
             print!("{}", ctx.content);
             if !ctx.content.ends_with('\n') {
@@ -178,12 +198,13 @@ pub fn show(workspace_id: &str, name: &str) {
 }
 
 pub fn pull(workspace_id: &str, name: &str, force: bool, dry_run: bool) {
-    if let Err(e) = validate_context_stem(name) {
+    let name = normalize_context_cli_name(name);
+    if let Err(e) = validate_context_stem(&name) {
         eprintln!("error: {e}");
         std::process::exit(1);
     }
 
-    let path = local_md_path(name);
+    let path = local_md_path(&name);
 
     if !dry_run && !force && path.exists() {
         eprintln!(
@@ -194,7 +215,7 @@ pub fn pull(workspace_id: &str, name: &str, force: bool, dry_run: bool) {
     }
 
     let api = ApiClient::new(Some(workspace_id));
-    let ctx = match fetch_context(&api, name) {
+    let ctx = match fetch_context(&api, &name) {
         Ok(c) => c,
         Err(reqwest::StatusCode::NOT_FOUND) => {
             eprintln!(
@@ -232,12 +253,13 @@ pub fn pull(workspace_id: &str, name: &str, force: bool, dry_run: bool) {
 }
 
 pub fn push(workspace_id: &str, name: &str, dry_run: bool) {
-    if let Err(e) = validate_context_stem(name) {
+    let name = normalize_context_cli_name(name);
+    if let Err(e) = validate_context_stem(&name) {
         eprintln!("error: {e}");
         std::process::exit(1);
     }
 
-    let path = local_md_path(name);
+    let path = local_md_path(&name);
     if !path.is_file() {
         eprintln!(
             "{}",
@@ -269,7 +291,7 @@ pub fn push(workspace_id: &str, name: &str, dry_run: bool) {
     }
 
     let api = ApiClient::new(Some(workspace_id));
-    let body = json!({ "name": name, "content": content });
+    let body = json!({ "name": &name, "content": content });
     let resp: UpsertResponse = api.post("/context", &body);
 
     println!(
@@ -330,4 +352,26 @@ mod tests {
     fn validate_rejects_reserved_uppercase() {
         assert!(validate_context_stem("SELECT").is_err());
     }
+
+    #[test]
+    fn normalize_strips_trailing_md() {
+        assert_eq!(normalize_context_cli_name("USER.md"), "USER");
+        assert_eq!(normalize_context_cli_name("USER.MD"), "USER");
+        assert_eq!(normalize_context_cli_name("  USER.md  "), "USER");
+    }
+
+    #[test]
+    fn normalize_accepts_path_with_md() {
+        assert_eq!(normalize_context_cli_name("./DATAMODEL.md"), "DATAMODEL");
+    }
+
+    #[test]
+    fn normalize_preserves_stem_without_md() {
+        assert_eq!(normalize_context_cli_name("DATAMODEL"), "DATAMODEL");
+    }
+
+    #[test]
+    fn normalize_strips_md_one_char_stem() {
+        assert_eq!(normalize_context_cli_name("a.md"), "a");
+    }
 }