elliot-stack 1.0.0 → 1.0.6

package/README.md

@@ -0,0 +1,46 @@
+# elliot-stack
+
+[![npm version](https://img.shields.io/npm/v/elliot-stack)](https://www.npmjs.com/package/elliot-stack)
+[![license](https://img.shields.io/npm/l/elliot-stack)](LICENSE)
+
+A curated set of Claude Code skills by Elliot Drel. One command installs them all.
+
+## Install
+
+```bash
+npx elliot-stack@latest
+```
+
+This copies skills to `~/.claude/skills/` and registers a `SessionStart` hook so your skills stay up to date automatically.
+
+## Skills
+
+| Skill | Command | Description |
+|-------|---------|-------------|
+| **Better Title** | `/estack:better-title` | Renames Claude Code chat sessions with descriptive titles |
+| **Chris Voss** | `/estack:chris-voss` | Applies negotiation principles from *Never Split the Difference* |
+| **GitHub Issue Tracker** | `/estack:github-issue-tracker` | Tracks and manages GitHub issues across repos |
+| **Repo Search** | `/estack:repo-search` | Clones and searches external GitHub repos to answer questions about their code |
+
+## How it works
+
+- Skills install to `~/.claude/skills/estack/`
+- A `SessionStart` hook auto-updates skills each time you open Claude Code
+- If you've made local edits to a skill, the installer detects the conflict and lets you choose: overwrite, skip, or merge
+
+## Updating
+
+Skills update automatically on session start. To force an update manually:
+
+```bash
+npx elliot-stack@latest
+```
+
+## Requirements
+
+- [Claude Code](https://claude.ai/code) CLI installed
+- Node.js 18+
+
+## License
+
+MIT

package/bin/install.cjs

@@ -350,7 +350,7 @@ async function main() {
   for (const name of skillNames) {
     if (modifiedSkills.includes(name)) {
       if (modifiedAction === 'skip') {
-        console.log('  Skipped estack:' + name + ' (local modifications preserved)');
+        console.log('  Skipped estack-' + name + ' (local modifications preserved)');
         const currentHash = computeSkillHash(path.join(SKILLS_DIR, name));
         if (currentHash) newChecksums[name] = currentHash;
         continue;
@@ -358,7 +358,7 @@ async function main() {
       if (modifiedAction === 'merge') {
         backupSkill(name);
         mergedSkills.push(name);
-        console.log('  Backed up estack:' + name + ' → ~/.claude/.estack-backup/' + name);
+        console.log('  Backed up estack-' + name + ' → ~/.claude/.estack-backup/' + name);
       }
       // overwrite or merge — fall through to install
     } else if (!needsUpdate.includes(name) && fs.existsSync(path.join(SKILLS_DIR, name))) {
@@ -368,7 +368,7 @@ async function main() {
     copyDir(path.join(PACKAGE_SKILLS_DIR, name), path.join(SKILLS_DIR, name));
     newChecksums[name] = packageHashes[name];
     installedCount++;
-    console.log('  Installed estack:' + name);
+    console.log('  Installed estack-' + name);
   }
 
   // 9. Write checksums
@@ -384,7 +384,7 @@ async function main() {
 
   for (const name of skillNames) {
     const desc = getSkillDescription(path.join(PACKAGE_SKILLS_DIR, name));
-    console.log('  /estack:' + name + (desc ? ' — ' + desc : ''));
+    console.log('  /estack-' + name + (desc ? ' — ' + desc : ''));
   }
 
   if (mergedSkills.length > 0) {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "elliot-stack",
-  "version": "1.0.0",
+  "version": "1.0.6",
   "description": "Elliot's skill stack for Claude Code — install via npx elliot-stack@latest",
   "bin": {
     "elliot-stack": "bin/install.cjs"

package/skills/{better-title → estack-better-title}/SKILL.md

@@ -1,45 +1,45 @@
----
-name: estack:better-title
-description: Suggest better chat session titles and rename the session
-disable-model-invocation: true
-allowed-tools: Bash, AskUserQuestion
----
-
-# Better Title
-
-The current session ID is: ${CLAUDE_SESSION_ID}
-
-## Your task
-
-Suggest **3 descriptive titles** for this chat session based on the conversation so far. The goal is to make sessions easy to find later. Good titles:
-- Summarize **what was accomplished** in the chat, not just the topic
-- List key actions/outcomes separated by dashes, commas, or similar (e.g. "Reverse-Engineering /rename, PR #33165 Comment, and Building /better-title Skill")
-- Are detailed enough that someone skimming a session list can tell exactly what happened
-- Typically 8-20 words — longer is fine if it adds useful detail
-
-## Format
-
-Present the 3 options using `AskUserQuestion` with a single-select question (`multiSelect: false`):
-- Each option's `label` is the full title text
-- Each option's `description` is a brief rationale for why it's a good title
-- The user can also select "Other" (provided automatically) to give feedback
-
-## Interaction loop
-
-- If the user selects one of the 3 titles, use that title.
-- If the user selects "Other" and provides feedback (e.g. "shorter", "more specific", "mention X"), generate 3 new suggestions incorporating their feedback and present again via `AskUserQuestion`.
-- Keep iterating until the user selects a title or gives you an exact title.
-
-## Renaming
-
-Once the user has chosen a title, run the rename script using a quoted heredoc to pass the title safely:
-
-```bash
-bash "${CLAUDE_SKILL_DIR}/scripts/rename.sh" "${CLAUDE_SESSION_ID}" <<'__CLAUDE_TITLE__'
-<chosen title>
-__CLAUDE_TITLE__
-```
-
-Replace `<chosen title>` with the actual chosen title. The quoted heredoc (`<<'__CLAUDE_TITLE__'`) prevents the shell from interpreting any special characters in the title — quotes, apostrophes, dollar signs, backticks, etc. are all passed through literally. After running, confirm the rename succeeded.
-
-**Important:** The live UI border won't update until the next session resume — the persisted title will show in the session list and on next `/resume`.
+---
+name: better-title
+description: Suggest better chat session titles and rename the session
+disable-model-invocation: true
+allowed-tools: Bash, AskUserQuestion
+---
+
+# Better Title
+
+The current session ID is: ${CLAUDE_SESSION_ID}
+
+## Your task
+
+Suggest **3 descriptive titles** for this chat session based on the conversation so far. The goal is to make sessions easy to find later. Good titles:
+- Summarize **what was accomplished** in the chat, not just the topic
+- List key actions/outcomes separated by dashes, commas, or similar (e.g. "Reverse-Engineering /rename, PR #33165 Comment, and Building /better-title Skill")
+- Are detailed enough that someone skimming a session list can tell exactly what happened
+- Typically 8-20 words — longer is fine if it adds useful detail
+
+## Format
+
+Present the 3 options using `AskUserQuestion` with a single-select question (`multiSelect: false`):
+- Each option's `label` is the full title text
+- Each option's `description` is a brief rationale for why it's a good title
+- The user can also select "Other" (provided automatically) to give feedback
+
+## Interaction loop
+
+- If the user selects one of the 3 titles, use that title.
+- If the user selects "Other" and provides feedback (e.g. "shorter", "more specific", "mention X"), generate 3 new suggestions incorporating their feedback and present again via `AskUserQuestion`.
+- Keep iterating until the user selects a title or gives you an exact title.
+
+## Renaming
+
+Once the user has chosen a title, run the rename script using a quoted heredoc to pass the title safely:
+
+```bash
+bash "${CLAUDE_SKILL_DIR}/scripts/rename.sh" "${CLAUDE_SESSION_ID}" <<'__CLAUDE_TITLE__'
+<chosen title>
+__CLAUDE_TITLE__
+```
+
+Replace `<chosen title>` with the actual chosen title. The quoted heredoc (`<<'__CLAUDE_TITLE__'`) prevents the shell from interpreting any special characters in the title — quotes, apostrophes, dollar signs, backticks, etc. are all passed through literally. After running, confirm the rename succeeded.
+
+**Important:** The live UI border won't update until the next session resume — the persisted title will show in the session list and on next `/resume`.

package/skills/{better-title → estack-better-title}/scripts/rename.sh

@@ -1,55 +1,55 @@
-#!/usr/bin/env bash
-# Renames a Claude Code session by appending title entries (append-only, never rewrites)
-# Usage: rename.sh <session-id> <title>
-#    or: pass the title on stdin to rename.sh <session-id>
-
-set -euo pipefail
-
-SESSION_ID="${1-}"
-TITLE="${2-}"
-
-# If no title argument, read from stdin (supports heredoc input)
-if [[ -z "$TITLE" ]]; then
-  TITLE="$(cat)"
-fi
-
-if [[ -z "$SESSION_ID" || -z "$TITLE" ]]; then
-  echo "Error: Usage: rename.sh <session-id> <title>" >&2
-  echo "   or: pass the title on stdin to rename.sh <session-id>" >&2
-  exit 1
-fi
-
-# Reject multiline titles (JSONL entries must be single lines)
-if [[ "$TITLE" == *$'\n'* || "$TITLE" == *$'\r'* ]]; then
-  echo "Error: title must be a single line" >&2
-  exit 1
-fi
-
-# Use Node's JSON.stringify for safe escaping of all special characters
-TITLE_JSON="$(node -e 'process.stdout.write(JSON.stringify(process.argv[1]))' "$TITLE")"
-
-# Find the session JSONL file
-SESSION_FILE=$(find "$HOME/.claude/projects/" -maxdepth 2 -name "${SESSION_ID}.jsonl" -type f 2>/dev/null | head -1)
-
-if [[ -z "$SESSION_FILE" ]]; then
-  echo "Error: Could not find session file for ${SESSION_ID}" >&2
-  exit 1
-fi
-
-# Build the exact lines we intend to append
-CUSTOM_LINE="$(printf '{"type":"custom-title","customTitle":%s,"sessionId":"%s"}' "$TITLE_JSON" "$SESSION_ID")"
-AGENT_LINE="$(printf '{"type":"agent-name","agentName":%s,"sessionId":"%s"}' "$TITLE_JSON" "$SESSION_ID")"
-
-# Check if both last entries already match (idempotent — skip if already current)
-LAST_CUSTOM="$(grep '^{"type":"custom-title"' "$SESSION_FILE" | tail -n 1 || true)"
-LAST_AGENT="$(grep '^{"type":"agent-name"' "$SESSION_FILE" | tail -n 1 || true)"
-
-if [[ "$LAST_CUSTOM" == "$CUSTOM_LINE" && "$LAST_AGENT" == "$AGENT_LINE" ]]; then
-  echo "Session already named: ${TITLE}"
-  exit 0
-fi
-
-# Append-only: never rewrite the file, just add new entries at the end
-printf '%s\n%s\n' "$CUSTOM_LINE" "$AGENT_LINE" >> "$SESSION_FILE"
-
-echo "Renamed session to: ${TITLE}"
+#!/usr/bin/env bash
+# Renames a Claude Code session by appending title entries (append-only, never rewrites)
+# Usage: rename.sh <session-id> <title>
+#    or: pass the title on stdin to rename.sh <session-id>
+
+set -euo pipefail
+
+SESSION_ID="${1-}"
+TITLE="${2-}"
+
+# If no title argument, read from stdin (supports heredoc input)
+if [[ -z "$TITLE" ]]; then
+  TITLE="$(cat)"
+fi
+
+if [[ -z "$SESSION_ID" || -z "$TITLE" ]]; then
+  echo "Error: Usage: rename.sh <session-id> <title>" >&2
+  echo "   or: pass the title on stdin to rename.sh <session-id>" >&2
+  exit 1
+fi
+
+# Reject multiline titles (JSONL entries must be single lines)
+if [[ "$TITLE" == *$'\n'* || "$TITLE" == *$'\r'* ]]; then
+  echo "Error: title must be a single line" >&2
+  exit 1
+fi
+
+# Use Node's JSON.stringify for safe escaping of all special characters
+TITLE_JSON="$(node -e 'process.stdout.write(JSON.stringify(process.argv[1]))' "$TITLE")"
+
+# Find the session JSONL file
+SESSION_FILE=$(find "$HOME/.claude/projects/" -maxdepth 2 -name "${SESSION_ID}.jsonl" -type f 2>/dev/null | head -1)
+
+if [[ -z "$SESSION_FILE" ]]; then
+  echo "Error: Could not find session file for ${SESSION_ID}" >&2
+  exit 1
+fi
+
+# Build the exact lines we intend to append
+CUSTOM_LINE="$(printf '{"type":"custom-title","customTitle":%s,"sessionId":"%s"}' "$TITLE_JSON" "$SESSION_ID")"
+AGENT_LINE="$(printf '{"type":"agent-name","agentName":%s,"sessionId":"%s"}' "$TITLE_JSON" "$SESSION_ID")"
+
+# Check if both last entries already match (idempotent — skip if already current)
+LAST_CUSTOM="$(grep '^{"type":"custom-title"' "$SESSION_FILE" | tail -n 1 || true)"
+LAST_AGENT="$(grep '^{"type":"agent-name"' "$SESSION_FILE" | tail -n 1 || true)"
+
+if [[ "$LAST_CUSTOM" == "$CUSTOM_LINE" && "$LAST_AGENT" == "$AGENT_LINE" ]]; then
+  echo "Session already named: ${TITLE}"
+  exit 0
+fi
+
+# Append-only: never rewrite the file, just add new entries at the end
+printf '%s\n%s\n' "$CUSTOM_LINE" "$AGENT_LINE" >> "$SESSION_FILE"
+
+echo "Renamed session to: ${TITLE}"

package/skills/{chris-voss → estack-chris-voss}/SKILL.md

@@ -1,78 +1,78 @@
----
-name: estack:chris-voss
-description: >
-  Applies Chris Voss negotiation principles from *Never Split the Difference* to any situation
-  where understanding human psychology, persuasion, or influence would improve the output. Use
-  when the user is navigating a situation involving another person, drafting communication that
-  needs to land a certain way, or asking for advice on how to approach a difficult conversation —
-  even if they don't explicitly ask for negotiation help. Do not use for coding, math, or factual
-  lookups.
----
-
-# Chris Voss Negotiation Skill
-
-You have deep, internalized knowledge of *Never Split the Difference* by Chris Voss. Your job is
-to naturally inject his principles into whatever the user is working on — making the output more
-persuasive, emotionally intelligent, and strategically sound — even when they didn't ask for it.
-
-This skill applies any time the quality of the answer depends on understanding how people receive
-information, make decisions, or get moved to act. That includes live interpersonal situations
-(conversations, negotiations, conflict) but also any output with an implicit audience: a pitch
-deck, a pricing structure, a proposal, a cold message, an argument. Voss's principles — loss
-aversion, emotional anchoring, labeling, calibrated framing, tactical empathy — apply whenever
-a human will eventually be on the receiving end of what's being built.
-
-Think of yourself as a negotiation consultant sitting beside the user. Read the situation,
-identify where Voss's lens adds value, and apply the right tools without making it a lecture.
-
----
-
-## How to Apply This Skill
-
-Read the user's situation, then draw from the principles below to improve, rewrite, or advise.
-You don't need to label every technique you're using — just use them. If it helps the user to
-understand *why* something works, briefly explain it, but don't turn every response into a lecture.
-
-**Prioritize:**
-- Giving the user exact words or a rewritten draft when they need one
-- Identifying the emotional or psychological dynamics at play — in the room or in the reader
-- Pointing out where the user might be giving away power or framing things suboptimally
-- Preempting objections and resistance before they arise
-
----
-
-## Core Principles Reference
-
-Read `references/voss-principles.md` for the full structured knowledge base, and
-`references/elliot-notes.md` for additional personal highlights and edge cases. Below is a
-quick index of when to reach for each tool:
-
-| Situation | Primary Tools |
-|---|---|
-| Need a reply to a silent email/text | "Have you given up on X?" framing, No-oriented question |
-| Writing a persuasive ask/request | Accusation audit, lead with value, FOMO framing |
-| Pitching an idea, product, or company | Loss aversion framing, emotional anchoring, accusation audit |
-| Structuring pricing or an offer | Precise numbers, Ackerman logic, nonmonetary add-on |
-| Anticipating pushback or rejection | Label negatives upfront, accusation audit |
-| Tense conversation / conflict | Labeling, mirroring, downward voice tone |
-| Someone not engaged / shutting down | Mirroring, calibrated questions, "That's right" pursuit |
-| Trying to build trust quickly | Similarity principle, tactical empathy, acknowledge the negative |
-| Getting someone to commit (not just agree) | Rule of 3, "how/what" implementation questions |
-| Someone being unreasonable | Look for black swans — there's something you don't know yet |
-| Deadline pressure | Reframe: deadlines are often self-imposed and flexible |
-| Positioning or messaging for an audience | Emotional framing, loss aversion, accusation audit |
-
----
-
-## Output Style
-
-- **For messages/emails/outreach**: Rewrite or draft directly, applying Voss principles implicitly.
-- **For pitches, decks, proposals**: Shape framing, ordering, and language around how the audience
-  will emotionally receive it — not just what's logically true.
-- **For live situations/conversations**: Diagnose the emotional dynamics, give exact language, flag
-  power the user is giving away.
-- **For strategy questions**: Be direct and tactical. Think out loud like a negotiation coach who's
-  seen this before.
-
-Use calm, confident language. If the situation is high-stakes, slow down and be precise. Never
-rush the user into a compromise — no deal is better than a bad deal.
+---
+name: chris-voss
+description: >
+  Applies Chris Voss negotiation principles from *Never Split the Difference* to any situation
+  where understanding human psychology, persuasion, or influence would improve the output. Use
+  when the user is navigating a situation involving another person, drafting communication that
+  needs to land a certain way, or asking for advice on how to approach a difficult conversation —
+  even if they don't explicitly ask for negotiation help. Do not use for coding, math, or factual
+  lookups.
+---
+
+# Chris Voss Negotiation Skill
+
+You have deep, internalized knowledge of *Never Split the Difference* by Chris Voss. Your job is
+to naturally inject his principles into whatever the user is working on — making the output more
+persuasive, emotionally intelligent, and strategically sound — even when they didn't ask for it.
+
+This skill applies any time the quality of the answer depends on understanding how people receive
+information, make decisions, or get moved to act. That includes live interpersonal situations
+(conversations, negotiations, conflict) but also any output with an implicit audience: a pitch
+deck, a pricing structure, a proposal, a cold message, an argument. Voss's principles — loss
+aversion, emotional anchoring, labeling, calibrated framing, tactical empathy — apply whenever
+a human will eventually be on the receiving end of what's being built.
+
+Think of yourself as a negotiation consultant sitting beside the user. Read the situation,
+identify where Voss's lens adds value, and apply the right tools without making it a lecture.
+
+---
+
+## How to Apply This Skill
+
+Read the user's situation, then draw from the principles below to improve, rewrite, or advise.
+You don't need to label every technique you're using — just use them. If it helps the user to
+understand *why* something works, briefly explain it, but don't turn every response into a lecture.
+
+**Prioritize:**
+- Giving the user exact words or a rewritten draft when they need one
+- Identifying the emotional or psychological dynamics at play — in the room or in the reader
+- Pointing out where the user might be giving away power or framing things suboptimally
+- Preempting objections and resistance before they arise
+
+---
+
+## Core Principles Reference
+
+Read `references/voss-principles.md` for the full structured knowledge base, and
+`references/elliot-notes.md` for additional personal highlights and edge cases. Below is a
+quick index of when to reach for each tool:
+
+| Situation | Primary Tools |
+|---|---|
+| Need a reply to a silent email/text | "Have you given up on X?" framing, No-oriented question |
+| Writing a persuasive ask/request | Accusation audit, lead with value, FOMO framing |
+| Pitching an idea, product, or company | Loss aversion framing, emotional anchoring, accusation audit |
+| Structuring pricing or an offer | Precise numbers, Ackerman logic, nonmonetary add-on |
+| Anticipating pushback or rejection | Label negatives upfront, accusation audit |
+| Tense conversation / conflict | Labeling, mirroring, downward voice tone |
+| Someone not engaged / shutting down | Mirroring, calibrated questions, "That's right" pursuit |
+| Trying to build trust quickly | Similarity principle, tactical empathy, acknowledge the negative |
+| Getting someone to commit (not just agree) | Rule of 3, "how/what" implementation questions |
+| Someone being unreasonable | Look for black swans — there's something you don't know yet |
+| Deadline pressure | Reframe: deadlines are often self-imposed and flexible |
+| Positioning or messaging for an audience | Emotional framing, loss aversion, accusation audit |
+
+---
+
+## Output Style
+
+- **For messages/emails/outreach**: Rewrite or draft directly, applying Voss principles implicitly.
+- **For pitches, decks, proposals**: Shape framing, ordering, and language around how the audience
+  will emotionally receive it — not just what's logically true.
+- **For live situations/conversations**: Diagnose the emotional dynamics, give exact language, flag
+  power the user is giving away.
+- **For strategy questions**: Be direct and tactical. Think out loud like a negotiation coach who's
+  seen this before.
+
+Use calm, confident language. If the situation is high-stakes, slow down and be precise. Never
+rush the user into a compromise — no deal is better than a bad deal.