---
name: local-conversation-history
description: >-
  Entry point for local AI conversation history across Claude Code, Codex and Kimi CLI: routes each
  request to the reader or continuation skill that owns it and keeps the cross-provider inventory.
  Use when the platform is unknown or plural ("what was I working on", "we discussed this once") and
  for any Kimi CLI history. When platform and action are already clear, load that skill directly.
argument-hint: "[keywords | session-id | workspace-path]"
---

# Local Conversation History — router

This skill decides **which** skill runs. It does not parse history itself, does
not own commands for a single provider, and never re-implements what an executor
already does. If you find yourself explaining flags for one provider, you are in
the wrong skill — hand off and stop.

## Route by platform × action

Establish two things before routing: **which platform** the conversation lived
on, and whether the user wants **evidence** (what was said/done) or
**resumption** (take the work forward).

| Platform | Read evidence | Continue the work |
|---|---|---|
| Claude Code | `daymade-claude-code:read-claude-code-history` | `daymade-claude-code:continue-claude-code-work` |
| OpenAI Codex | `daymade-claude-code:read-codex-history` | `daymade-claude-code:continue-codex-work` |
| Kimi CLI | `read-claude-code-history`, with the Kimi scope named — see **Provider scope** | no continuation skill exists |

Resumption always follows a read. The continuation skills require a verified
read receipt; routing straight to them without one is a defect, not a shortcut.

**When the platform is not stated** — a bare session ID, "pick up where we left
off" — do not guess it. Identify it first: try the Claude Code exact-session
lookup in `read-claude-code-history`, then the Codex rollout locator in
`read-codex-history`. Only a lookup that returns a verified identity decides
which continuation skill runs; a plausible-looking ID prefix does not.

## Provider scope — the job only this entry point routes

Each executor defaults to its own provider, so a request that spans providers
never widens by itself. **Naming the scope is this skill's whole job.** It has
two axes, and they use different flags — conflating them is the failure this
section exists to prevent.

| Cross-provider need | Route to | Name this scope |
|---|---|---|
| **Inventory** — "what have I been working on", "list my recent chats", session titles/dates/IDs | `read-claude-code-history`, its bundled inventory | `--source all`, or `--source kimi` for Kimi alone |
| **Content search** — "did we ever discuss X", find the conversation containing a quote, file, or tool result | `read-claude-code-history`, its bundled full-event search | add `--codex` and `--kimi` to the Claude search; each is a separate store the Claude registry never covers |
| **Ranked recall** — the same question when the wording may have drifted, or the sweep has no session ID, date, or project to bound it | `read-claude-code-history`, its optional hybrid recall index | The index states which providers it holds; read the coverage line it prints instead of assuming it spans all three. When the platform may also have been Codex, the Claude hybrid index does not cover that store — add `read-codex-history`'s `claude-flow-viewer` full-text path. That path matches literally only, so a genuine paraphrase still needs the Claude-side hybrid: the two are complementary, not substitutes |

Both readers ship the same inventory command and its `--source` already defaults
to `all` — but each reader's own task table pins it to that reader's provider
(`--source claude`, `--source codex`), so the default never fires on its own.
Search is the mirror image: it is Claude-only unless the other two stores are
added explicitly.

**Use supplied clues before broad discovery.** For a pasted quote with a known
project, date, or title, let the owning reader's inventory narrow candidates, then
verify their original messages. For Codex, follow **Locate a quoted exchange** in
`read-codex-history`. Inventory alone never establishes a content match; if its
candidates miss or its scope is incomplete, widen through the content-search row.
Keep the current Session excluded. A request for only an ID stops at verified
message evidence; it does not require reconstructing every unrelated conversation.

For this single-ID lookup when the provider is unknown, first probe the Codex
reader's inventory with `--source codex`: it uses state-DB metadata when available.
This is a discovery order, not an assumption that the conversation was Codex;
only verified original messages establish that. If no candidate verifies, widen
to the other providers through the content-search row. Do not start this probe
with `--source all`: its Claude inventory reads session bodies before applying
date and output limits. Complete cross-provider inventories, exhaustive searches,
and absence claims still require their full requested coverage.

**Without useful bounding clues, order the last two rows rather than picking one.** A cross-provider content
search is the expensive shape: it reads every event of every store, so the cost
scales with the whole corpus rather than with the question. When an index exists
and covers the providers in scope, recall answers in about a second and returns
leads — sessions, dates, projects — that turn the exhaustive scan into a bounded
one. Run it first, then scan only what the index does not hold. Skip straight to
search when the request needs an exhaustive guarantee, because ranked recall
returns top-K candidates and can never support an absence claim.

The index is optional. On a machine that never built one, recall exits non-zero
saying the index does not exist — that is a routing signal, not a failure to
report: fall back to the search row and say the sweep ran unindexed.

**Kimi CLI has no other entry anywhere** — no dedicated skill exists for either
axis, so both routes above land in `read-claude-code-history`, which documents
its own Kimi home resolution.

Let the executor own every flag beyond provider scope: `--all-projects`,
`--recursive`, date bounds, `--include-archived`, `--include-subagents`,
`--include-automated`, output format, and every detail of how each store is
parsed. This skill names which providers are in scope and nothing else.

## Intent decides the route — the word "history" does not

| The user's requested result | Route |
|---|---|
| A list of conversations: titles, dates, session IDs | The inventory row under **Provider scope**, or the matching reader when one platform is named |
| The conversation where a topic, quote, file, or tool result appeared — "find that old chat", "did we ever discuss X" | The **search** row under **Provider scope**; use supplied clues to narrow candidates first. Listing titles alone is not searching content, and a title match is not evidence the content exists |
| Their own raw inputs in chronological order, verbatim | The matching reader's verbatim-input path. Preserve duplicates and session boundaries; duplicates are part of the ledger, not noise |
| Picking work back up from an identified session | The matching continuation skill, after a read |

The requested output wins over the background motivation. If someone explains a
problem and then asks for a window of their own raw inputs, return that window —
the explanation's topic clues do not convert the request into a content search.

## Invariants that survive routing

- **Completeness.** A Claude inventory's source set is indivisible: the
  auto-discovered active homes (`~/.claude`, profile homes, the current
  `CLAUDE_CONFIG_DIR`) **plus** every archive registered in
  `~/.claude/history-sources.json`. Never call a conversation absent unless the
  output shows the registered archives were covered. An unavailable required
  archive is a configuration error, not permission to return a partial answer.
  `--claude-home` is a diagnostic override and can never back a completeness
  claim.
- **Self-match.** The current session records the user's question and this
  agent's own commands, so it matches almost any query about itself. Exclude the
  current session ID before treating a hit as historical evidence.
- **Zero results are not absence.** Ranked recall and a bounded search both
  return nothing for wording that exists under different words. Widen, or say
  what was searched — do not convert an empty result into "it never happened".
- **Ask whether a word is the topic's name or the way you talk about it.** A
  search for `技术选型` over a corpus where the user actually said 用哪个 and
  不要闭门造车 returns almost nothing — measured: 10 hits against 163 real
  occurrences, because a Chinese phrase only matches on a token boundary. When
  searching one person's corpus, search the way they speak: imperatives,
  negations, and concrete scene sentences, not the label a design document would
  give the topic. A zero hit on a topic name cannot support an absence claim;
  widen to the phrasing that person would have used first, and say which forms
  you tried.
- **A zero has three causes and only one of them is "no history".** The other
  two are a home that was never found and a scope that excluded everything, and
  none of the three looks different in an empty table. This bites Kimi hardest:
  its documented default home is not where every install puts it — a desktop
  client can bundle the CLI inside its own runtime and keep sessions there — and
  its sessions belong to their own workspaces, so a default run inside some other
  repository returns nothing on a store full of conversations. Rule out both
  before reporting absence: name the home that was actually read, and say what
  project scope was in effect. The inventory prints a diagnostic line when a home
  is missing, so quote it when it appears — but a **located** home that yields
  zero prints no diagnostic at all, and the search path prints none in either
  case, so never treat a silent empty result as the reader confirming absence.
  `read-claude-code-history` owns how to locate a home and which scope flags the
  inventory needs.

## Do not

- Do not run provider-specific parsing, SQLite, `rg`, `jq`, or JSONL pipelines
  here. Every one of those belongs to an executor that already handles its
  store's schema, archives, and failure modes.
- Do not copy an executor's flags into this file beyond the three that name
  provider scope (`--source`, `--codex`, `--kimi`) — those are this skill's own
  subject. Every other flag changes on the executor's schedule; copying one here
  makes this file drift silently and then teach the wrong command.
- Do not route to a continuation skill to answer a question about the past.
  Reading is evidence; continuing changes the world.