refactor(edit): removed unsupported ignore-case and multiline sed flags

- Removed `i` and `m` from the Atom sed schema and docs, so sed now accepts only `pat`, `rep`, `g`, and `F` options.
- Updated `SedSpec` parsing and serialization to drop ignore-case and multiline handling.
- Adjusted regex construction to apply sed replacements using only global and literal flags.

packages/coding-agent/src/edit/modes/atom.ts

@@ -12,7 +12,7 @@
  *   { path, loc: "5th",      pre: [...], splice: [...], post: [...] }
  *   { path, loc: "$",        pre:  [...] }                            // prepend to file
  *   { path, loc: "$",        post: [...] }                            // append to file
- *   { path, loc: "$",        sed:  { pat, rep, g?, F?, i?, m? } }         // sed on every line
+ *   { path, loc: "$",        sed:  { pat, rep, g?, F? } }                 // sed on every line
  *
  * `splice: []` on a single-anchor locator deletes that line. `splice:[""]` preserves
  * a blank line. Line ranges are not supported.
@@ -72,8 +72,6 @@ export const atomEditSchema = Type.Object(
 					rep: Type.String({ description: "replacement text" }),
 					g: Type.Optional(Type.Boolean({ description: "global replace", default: false })),
 					F: Type.Optional(Type.Boolean({ description: "literal replace", default: false })),
-					i: Type.Optional(Type.Boolean({ description: "ignore case", default: false })),
-					m: Type.Optional(Type.Boolean({ description: "multiline mode (^ and $ match line boundaries)", default: false })),
 				},
 				{
 					additionalProperties: false,
@@ -110,11 +108,9 @@ export type AtomEdit =
 	| { op: "sed_file"; spec: SedSpec; expression: string };
 
 export interface SedSpec {
-	multiline: boolean;
 	pattern: string;
 	replacement: string;
 	global: boolean;
-	ignoreCase: boolean;
 	literal: boolean;
 }
 
@@ -270,7 +266,7 @@ function extractAnchorContentHint(raw: string): string | undefined {
 
 function parseSedSpec(input: unknown, editIndex: number): SedSpec {
 	if (input === null || typeof input !== "object" || Array.isArray(input)) {
-		throw new Error(`Edit ${editIndex}: sed must be an object with shape {pat, rep, g?, F?, i?, m?}.`);
+		throw new Error(`Edit ${editIndex}: sed must be an object with shape {pat, rep, g?, F?}.`);
 	}
 	const obj = input as Record<string, unknown>;
 	const pat = obj.pat;
@@ -286,7 +282,7 @@ function parseSedSpec(input: unknown, editIndex: number): SedSpec {
 	if (typeof rep !== "string") {
 		throw new Error(`Edit ${editIndex}: sed.rep must be a string.`);
 	}
-	const readBool = (key: "g" | "F" | "i" | "m", defaultValue: boolean): boolean => {
+	const readBool = (key: "g" | "F", defaultValue: boolean): boolean => {
 		const v = obj[key];
 		if (v === undefined) return defaultValue;
 		if (typeof v !== "boolean") {
@@ -296,21 +292,17 @@ function parseSedSpec(input: unknown, editIndex: number): SedSpec {
 	};
 	const global = readBool("g", false);
 	const literal = readBool("F", false);
-	const ignoreCase = readBool("i", false);
-	const multiline = readBool("m", false);
-	return { pattern: pat, replacement: rep, global, literal, ignoreCase, multiline };
+	return { pattern: pat, replacement: rep, global, literal };
 }
 
 function formatSedExpression(spec: SedSpec): string {
-	const obj: { pat: string; rep: string; g?: boolean; F?: boolean; i?: boolean; m?: boolean } = {
+	const obj: { pat: string; rep: string; g?: boolean; F?: boolean } = {
 		pat: spec.pattern,
 		rep: spec.replacement,
 	};
 	// Only emit non-default flags so error messages stay compact (g defaults false).
 	if (spec.global) obj.g = true;
 	if (spec.literal) obj.F = true;
-	if (spec.ignoreCase) obj.i = true;
-	if (spec.multiline) obj.m = true;
 	return JSON.stringify(obj);
 }
 
@@ -335,8 +327,6 @@ function applySedToLine(
 	}
 	let flags = "";
 	if (spec.global) flags += "g";
-	if (spec.ignoreCase) flags += "i";
-	if (spec.multiline) flags += "m";
 	let re: RegExp | undefined;
 	let compileError: string | undefined;
 	try {

packages/coding-agent/src/prompts/tools/atom.md

@@ -14,13 +14,11 @@ Verbs:
 - `splice: […]`: lines are spliced in at the anchor.
 - `pre: […]`: prepend before the anchor (or at BOF if `loc=$`)
 - `post: […]`: append after the anchor (or at EOF if `loc=$`)
-- `sed: { pat, rep, g?, F?, i?, m? }` — structured find/replace on the anchor line. **Prefer this over `splice` for token-level changes**
+- `sed: { pat, rep, g?, F? }` — structured find/replace on the anchor line. **Prefer this over `splice` for token-level changes**
   - `pat`: pattern to find (regex by default)
   - `rep`: replacement (regex back-refs like `$1`, `$&` available)
   - `g`: global — replace every occurrence (default `false`; pass `true` to replace all)
   - `F`: literal — treat `pat` as a literal substring (no regex). Use this whenever `pat` contains `||`, `.`, `(`, `?`, `\`, etc. you mean literally.
-  - `i`: ignore case
-  - `m`: multiline — `^` and `$` match line boundaries
 You **MUST** keep `pat` as short as possible.
 
 Combination rules: